19
No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020
1 You can ind his ex and i s DOI a h ps:// idssk i .dk/njwls/index.
2 Co esponding au ho : Mii a Niska, Facul y o Social Sciences (Uni o Social Resea ch), 33014
Tampe e Uni e si y, Finland. E-mail: [email p o ec ed]. ‡ Con ibu ed equally.
E alua ion o Li elong Wo k Sa is ac ion among Fo me
Pos al Se ice Wo ke s1
❚❚ Mii a Niska2‡
Pos doc o al esea che , Facul y o Social Sciences (Uni o Social Resea ch), Tampe e Uni e si y,
Tampe e, Finland
❚❚ P akash, K. C.‡
Pos doc o al esea che , Facul y o Social Sciences (Uni o Heal h Sciences), Ge on ology
Resea ch Cen e , Tampe e Uni e si y, Finland
❚❚ Anna Siukola
Pos doc o al esea che , Facul y o Social Sciences (Uni o Heal h Sciences), Ge on ology
Resea ch Cen e , Tampe e Uni e si y, Finland
❚❚ Hanna Kosonen
Resea ch assis an , Facul y o Social Sciences (Uni o Heal h Sciences), Tampe e Uni e si y,
Finland
❚❚ Ja i Luomanen
Pos doc o al esea che , Facul y o Social Sciences (Uni o Social Resea ch), Tampe e Uni e si y,
Finland
❚❚ Ki si Lumme-Sand
Uni e si y lec u e , Facul y o Social Sciences (Uni o Heal h Sciences), Ge on ology Resea ch
Cen e , Tampe e Uni e si y, Finland
❚❚ Subas Neupane
Pos doc o al esea che , Facul y o Social Sciences (Uni o Heal h Sciences), Ge on ology
Resea ch Cen e , Tampe e Uni e si y, Finland
❚❚ Pi jo Nikande
Resea ch di ec o , The Doc o al School, Tampe e Uni e si y, Finland
❚❚ Clas-Håkan Nygå d
P o esso , Facul y o Social Sciences (Uni o Heal h Sciences), Ge on ology Resea ch Cen e ,
Tampe e Uni e si y, Finland
20 E alua ion o Li elong Wo k Sa is ac ion Mii a Niska e al.
ABSTRACT
While policymake s emphasize he need o delay e i emen age, nume ous adi ional indus ies,
among hem pos al se ices, emphasize he need o employee downsizings. These can be con a-
dic o y needs. Downsizings can educe wo k li e sa is ac ion (WLS) among wo ke s, and educed
WLS can dec ease willingness o delay e i emen age. In his c oss-sec ional s udy, we explo e
WLS ajec o ies o o me pos al se ice wo ke s by analyzing quan i a i e su ey da a (N = 201)
and quali a i e in e iew da a (N = 40). In he da a, wo ke s aged 50–67 yea s e alua e hei
li elong wo k sa is ac ion. The esul s o he quan i a i e and he quali a i e s udy complemen and
mi o each o he . Same ajec o ies o WLS we e iden i ied in bo h da a. One- hi d o he su ey
esponden s epo ed dec easing WLS, which acco ding o he quali a i e da a ela es o employee
downsizings. The a icle highligh s he impo ance o acknowledging con ex ual elemen s o wo k
sa is ac ion measu emen and he bene i s o combining me hodological adi ions o unde s and
he longe - e m dynamics in ol ed.
KEYWORDS
Employee downsizing / ex ended wo king li e / li eline ajec o ies / mixing me hods / olde wo ke s /
wo k li e sa is ac ion
In oduc ion
Recen demog aphic shi s ha e igge ed a s ong poli ical need o delay e i emen
age and ex end wo king li es. O e he pas en yea s, he popula ion o No dic
coun ies has aged e en as e han in many o he Eu opean coun ies (G un elde
e al. 2018a). As he old age dependency a io inc eases, pushing he a e age e i emen
age u he is an appealing p oposi ion when compa ed agains aising axes o cu -
ing pensions (Eläke u akeskus 2013; Eu opean Commission 2010; Hess 2017; OECD
2017; Wise 2010). A he same ime, he labo ma ke is seeing majo shi s. In he
u u e, au oma ion and digi aliza ion can ake o e as much as 40 pe cen o wo k-
ing hou s in No dic coun ies (G un elde e al. 2018b). Toge he wi h globaliza ion,
au oma ion and digi aliza ion ha e o ced nume ous adi ional indus ies o in oke
downsizing s a egies ha include, o example, ea ly e i emen incen i es and layo s
(G un elde e al. 2018b; Macky 2004; McMullin & Ma shall 2001).
P e ious s udies ha e epo ed he associa ion o downsizing p ac ices wi h nega i e
psychological esponses like educed wo k sa is ac ion (Da a e al. 2010). Acco ding o
he classic de ini ion, wo k sa is ac ion is a posi i e emo ional s a e ha esul s om one’s
app aisal o one’s wo k (Locke 1976: 1300; also Spec o 1997). In his a icle, ou ocus is
on wo k li e sa is ac ion (WLS). Wi h his concep , we e e o wo k sa is ac ion h ough-
ou li e. Wo k sa is ac ion is i al o indi iduals’ wellbeing and happiness (Bojanowska
& Zalewska 2016), bu i is also i al o companies and success ul ex ension o wo king
li es: low wo k sa is ac ion associa es wi h e i emen in en ions and absen eeism om
wo k (Ha konmäki e al. 2009; Kade o s e al. 2016; Sagie 1998). I downsizing p ac ices
lowe wo k sa is ac ion, hey may hinde he ex ension o wo k ca ee s in wo ways:
olde ic ims ace di icul ies in inding new employmen (A ms ong-S assen & Ca aneo
2010; Fu unes & Mykle un 2010; Solem 2015), and olde su i o s s a hinking abou
ea ly e i emen (Bad an & Ka a y 2008; an den Be g e al. 2010).
Wo k sa is ac ion has been he opic o ac i e scien i ic inqui y (Be glund e al.
2017; Bussing e al. 1999; Kade o s e al. 2016; Pagán 2013). P e ious quan i a i e
No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020 21
s udies demons a e ha wo k- ela ed esou ces (S ynen e al. 2017), job suppo (Pagán
2013) and job con ol (Allan e al. 2018; Chao e al. 2015; Lind o s e al. 2007; Meye s
e al. 2019; Waddimba e al. 2016) a e posi i e p edic o s o wo k sa is ac ion. Simila ly,
wo k engagemen (Meye s e al. 2019; Shimazu & Schau eli 2009) and wo k s a us
(Na a o & Sal e da 2019; Tay & Ha e 2013; Vi anen e al. 2003) play a signi ican
ole in wo k sa is ac ion o wo ke s in a ious sec o s. In addi ion, poo pe cei ed heal h
(Lunau e al. 2014; Mäkikangas & Kinnunen 2003), inc eased numbe o sick lea e
days (Kuoppala e al. 2011), o ganiza ional injus ice (Lind o s e al. 2007; Meie e al.
2015; Pagán 2013) and i egula wo king hou s (Lunau e al. 2014) ha e he po en-
ial o deg ade wo k sa is ac ion. Likewise, poo quali y o wo k en i onmen (Sieg is
e al. 2007), employee age, hei enu e pe iod a wo k and wo king le el, ha is, he p o-
essional le el o place in he wo kplace/company hie a chy (Ramoo e al. 2013) could
educe wellbeing a wo k and ul ima ely esul in p ema u e exi om he labo ma ke .
Quan i a i e s udies on wo k sa is ac ion ha e been c i icized o dis ega ding he
pa icula his o ical and social con ex and na u e o indi iduals’ wo k sa is ac ion
e alua ions (Zembylas & Papanas asiou 2006). Whe eas Bojanowska and Zalewska
(2016) call o s udies on lay concep ions o happiness, Zembylas and Papanas asiou
(2006) call o s udies on lay concep ions o wo k sa is ac ion. Resea che s ha e also
conduc ed a subs an ial amoun o quali a i e s udies on wo k sa is ac ion (A e i e al.
2014; Bussing e al. 1999; de Wesley & Clemson 1992; McNeese-Smi h 1999; Moo e
e al. 2006). These indi idual in e iew and ocus g oup based s udies ha e mapped
he lay concep ions o wo k sa is ac ion, o example, o eache s in Cyp us, nu ses in
I an and occupa ional he apis s in Aus alia. In line wi h quan i a i e s udies, quali a-
i e s udies ha e demons a ed ha social ela ions, wo king condi ions and esou ces
ha e an e ec on wo k sa is ac ion (A e i e al. 2014; McNeese-Smi h 1999; Moo e
e al. 2006; Zembylas & Papanas asiou 2006). Fu he mo e, quali a i e s udies ha e
disclosed con ex -speci ic concep ions o wo k sa is ac ion. I anian nu ses, o example,
epo ed ha I anian, doc o -cen e ed cul u e dec eases hei wo k sa is ac ion (A e i
e al. 2014). Cyp io eache s epo ed ha he cen aliza ion o he Cyp io educa ional
sys em dec eases hei wo k sa is ac ion (Zembylas & Papanas asiou 2006). Conside -
ing hese con ex -speci ic concep ions o wo k sa is ac ion, i migh be insu icien o
measu e wo k sa is ac ion solely in quan i a i e e ms, and hus o e look he possibili-
ies o quali a i e esea ch (Bussing e al. 1999).
In his a icle, we s udy e alua ions o li elong wo k sa is ac ion o Finnish pos al
se ice wo ke s. The pos al se ice sec o all o e he wo ld aces massi e ans o ma-
ions due o p og essi e digi aliza ion and diminishing amoun o adi ional pape mail
(Dieke e al. 2013). In Finland, he s a e-owned company Pos i G oup has unde gone
signi ican s uc u al and ope a ional changes, such as company acquisi ions, me ge s,
selling pa s o he company, and enewals in se ices and ope a ions. Pos i G oup,
o me ly I ella and Suomen Pos i, used o ha e a epu a ion as a s able and eliable
employe (Pie iäinen 1998). Pos i G oup has also been one o he la ges co po a e
employe s in Finland. Howe e , o e he las decades, Pos i G oup has been laying o
i s pe sonnel. Finland has a law abou coope a ion in companies (334/2007), which p o-
mo es in e ac ional coope a ion p ocedu es be ween employe s and employees. When
companies wi h mo e han 20 employees make changes ha a ec hese employees,
such as downsizings, he employe is legally obliga ed o nego ia e wi h he employees
abou he easons, impac s and al e na i es o hose changes. In case o dismissals and
22 E alua ion o Li elong Wo k Sa is ac ion Mii a Niska e al.
layo s, he company is also esponsible o making a mo ion o employees’ e-employ-
men p omo ion, po en ially in collabo a ion wi h public employmen agencies. Since
2011, Pos i G oup has unde aken se e al coope a ion nego ia ions yea ly (Pos i annual
epo s 2011–2016). A e he coope a ion nego ia ions o 2014, a ound 780 people
we e laid o (Pos i annual epo 2014, p. 150). P e ious esea ch has demons a ed
ha a ai downsizing p ocess can ac ually mi iga e nega i e employee esponses such as
educing wo k sa is ac ion (Da a e al. 2010; Macky 2004). Has Pos i G oup been able
o conduc downsizings in a way ha p e en s educ ions in i s employees’ wo k sa is-
ac ion? The pu pose o his esea ch is o inc ease knowledge abou he ajec o ies o
li elong wo k sa is ac ion in he con ex o company downsizing. The aim is o acili a e
decision-making and hope ully inspi e u he esea ch in he ield o wo k sa is ac ion
s udies.
Wo k li e sa is ac ion is a signi ican conce n in e ms o he o e all heal h and
de elopmen o he indi idual wo ke (Lunau e al. 2014), and as such ca ies ac oss
impo an public heal h ou comes and e ec s. Due o massi e layo s, o me wo ke s o
he Pos i G oup ep esen a unique g oup o he s udy o WLS ansi ions in he con ex
o company downsizing. In o de o ind ou whe he cons an coope a ion nego ia-
ions and employee downsizings ha e educed employees’ WLS, we s udy WLS ajec o-
ies om bo h quan i a i e and quali a i e pe spec i es. Mos quan i a i e s udies ha e
a ge ed he WLS a a pa icula poin . Howe e , o map he possible changes, ou quan-
i a i e s udy in es iga es e ospec i ely he de elopmen al pa hways o WLS. The quan-
i a i e s udy also desc ibes he c oss-dis ibu ion o di e en basic and wo k- ela ed
cha ac e is ics o he s udy popula ion and hei membe ship in di e en de elopmen al
pa hways o WLS. To deepen ou unde s anding o he social con ex o WLS ajec o-
ies (see Spec o 1997), we also analyze na a i es, o s o ies, o empo al changes in
in e iewees’ subjec i e wo k sa is ac ion. We belie e ha his c oss- e iliza ion – o
mixing o wo di e en me hodologies – p o ides a ich desc ip ion o pos al wo ke s’
WLS ansi ions and hei causes.
Ma e ials and Me hods
To examine pos al wo ke s’ wo k sa is ac ion ajec o ies h oughou li e, we analyze
quan i a i e and quali a i e da a gene a ed in a esea ch p ojec Towa ds a wo-speed
Finland? (2015–2019). In he da a, o me pos al se ice wo ke s o e he age o 50
e alua e hei li elong wo k sa is ac ion.
The quan i a i e da a
Ini ially, a su ey ques ionnai e was sen o 4386 Finnish pos al se ice wo ke s and
44% o hem esponded. The selec ion o he laid-o esponden s o he c oss-sec ional
quan i a i e s udy is based on a su ey ques ionnai e, sen in 2018, o 1935 s udy sub-
jec s who we e ac i ely wo king o Pos i G oup un il 2016. A o al o 1466 esponded,
yielding a esponse a e o 76%. A ound 86% o he esponden s we e s ill wo king in
Pos i G oup and he es (n = 212) we e no longe wo king in Pos i G oup (laid-o ).
In he analysis, we s udied he laid-o esponden s who eplied o he ques ionnai e in
No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020 23
2016 ega ding hei WLS le el a 15–29 yea s, 50–59 yea s and in a leas one o he
in e ening li e ime poin s (ei he 30–39 yea s o 40–49 yea s o bo h). The inclusion
c i e ia esul ed in he selec ion o 201 esponden s.
Sa is ac ion in wo king li e (WLS)
The sa is ac ion le el o esponden s du ing he ages o ‘15–29’, ‘30–39’, ‘40–49’, ‘50–59’
and ‘60–69’ was asked using ques ionnai e. The esponses we e collec ed on a scale o ‘0’
( e y dissa is ied) – ‘10’ ( e y sa is ied). Since a la ge p opo ion o he esponden s we e
a ound 50–59 yea s o age, he ques ion ega ding hei sa is ac ion du ing ‘60–69’ yea s
o age was no included, which was a p oceeding ques ion and po en ial o in o ma ion
bias. The single i em o measu e WLS has been p e iously alida ed (Dolbie e al. 2005).
Wo k abili y
The i s i em o wo k abili y index ques ionnai e was used o assess he cu en wo k
abili y (Jääskeläinen e al. 2016). The esponden s we e asked o a e hei cu en wo k
abili y compa ed o li e’s bes using a scale o 0–10 whe e ‘0’ indica ed he wo s and
‘10’ indica ed he bes . The esponses we e ca ego ized in o Poo (0–5), Reasonable
(6 –7), Good (8–9) and Excellen (10). The ca ego ies Good and Excellen we e me ged
o be used as Good/Excellen wo k abili y in he p esen analysis. The Good and Excel-
len ca ego ies o wo k abili y we e me ged because he e we e e y minimal o no di -
e ences in ajec o y g oups wi h espec o hese wo ca ego ies, and a e me ging, he
single ca ego y was mo e ep esen a i e.
Job suppo and Job con ol
Job suppo was assessed wi h ques ions ela ed o eedback om supe iso s, suppo
om supe iso s and suppo om wo kma es, which we e answe ed on a scale o ‘0
o 10’, whe e ze o indica es low suppo and en indica es a high le el o suppo . The
summa y sco e o i e job suppo i ems (C onbach’s α = 0.78) anging om 8 o 50 was
dicho omized in o ‘high’ and ‘low’ using he median alue (36.0) ( on Bonsdo e al.
2012). Job con ol was assessed wi h ques ions ela ed o he esponden ’s possibili y o
lea n new knowledge and skills (0–10), possibili y o in luence wo k and wo king con-
di ions (1–4), an expe ience o doing impo an and signi ican wo k (1–6) and enough
educa ion a wo k (0–10) ( on Bonsdo e al. 2012). The summa y sco e o ou job
con ol i ems (C onbach’s α = 0.73) anging om 0 o 27 was dicho omized in o ‘high’
and ‘low’ using he median alue (17.0).
Co a ia es
The age o he esponden s anged om 51 o 67 yea s. The gende o he esponden s
was elucida ed as women and men. Two occupa ional classes, whi e colla (includes

24 E alua ion o Li elong Wo k Sa is ac ion Mii a Niska e al.
p o essional, manage ial and adminis a i e wo k) and blue colla (includes manual
labo and cus ome in e ac ion), we e used. The enu e pe iod in Pos i G oup was
assessed in a coun o yea s and wo king ime was assessed as wo king hou s pe week.
The numbe o sickness absence days pe yea was assessed in a coun o days and
is dicho omized in o ‘0–14’ and ‘15–365’ days o be used in his s udy. The le el o
educa ion is elucida ed in o h ee ca ego ies (basic schooling, college le el aining and
academic deg ee). The sel - epo ed esponses on he p esence o physician-diagnosed
disease we e assessed using yes/no ques ions. Likewise, equali y a wo k was assessed
using a yes/no ques ion. Pe cei ed heal h was assessed using a single i em ‘How do you
a e you cu en heal h compa ed o you li e’s bes ?’, and esponses we e collec ed on a
scale o ‘0–10’. Those esponses we e ca ego ized in o good (9–10), mode a e (7–8) and
poo / ai (0–6) o use in his s udy. The selec ion o he a iables was pa ly based on
p e ious li e a u e (Dolbie e al. 2005; Ramoo e al. 2013; Sieg is e al. 2007).
The quali a i e da a
The quali a i e da a we use o igina e om a mo e ex ensi e quali a i e esea ch p ojec
based on longi udinal quali a i e esea ch (e.g., Neale 2019). Fo he bene i s o his
a icle, we ocus on da a om he i s in e iew ound whe e in e iewees (N = 40)
e alua ed hei li elong wo k sa is ac ion ( o o e all da a gene a ed in he p ojec , see
able A1 in Appendix). In hese i s ound in e iews, esea ch pa icipan s we e asked
o d aw hei WLS li elines on a coo dina e plane whe e he x-axis mapped age (20,
30, 40, 50, 60, 70), and he y-axis mapped WLS (example o a li eline, see igu e A1
in Appendix). While d awing he li elines, in e iewees na a ed how hei wo k sa is-
ac ion had changed o e he yea s. A he ime o he i s in e iews, 20 pa icipan s
we e 50–59 yea s o age and 20 we e 60–65 yea s o age. Mos in e iewees (n = 17)
we e emale, blue-colla wo ke s. The o he in e iewees we e male blue-colla wo k-
e s (n = 8), emale whi e-colla wo ke s (n = 8) and male whi e-colla wo ke s (n = 7).
The in e iews we e eco ded and ansc ibed e ba im in Finnish. Fo he bene i s o
anspa ency and alidi y, we p esen he o iginal da a exce p ansc ip ions in Finnish
alongside hei ansla ions in o English. This is a commonly used means o secu ing he
alidi y and gua an eeing he publicly e i iable, anspa en na u e o he quali a i e
analysis (Nikande 2008).
Da a analysis
Mix u e modelling was used o de ec he de elopmen al pa hways o WLS. La en class
analysis (LCA) was used wi h con inuous la en class indica o s and use -speci ied s a -
ing alues based on he con inuous esponses o he sel - epo ed ques ion o WLS a
ou li e ime poin s in a c oss-sec ional da a. The esponses on wo king li e sa is ac-
ion a ages ‘15–29’, ‘30–39’, ‘40–49’ and ‘50–59’ we e used in he o iginal o m o
de ec he mos p obable de elopmen al pa hways e med as ajec o ies o sa is ac ion
in wo king li e in his s udy. LCA is a me hod ha iden i ies wi hin he da a, he mul iple
la en classes ha end o ha e a simila de elopmen o e ime (Mu hén 2003). We i -
ed he mix u e model wi h wo o ou classes and selec ed he bes - i ed model. The
No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020 25
selec ion o he bes - i ed model was based on Bayesian In o ma ion C i e ion (BIC),
Akaike In o ma ion C i e ion (AIC), subs an i e in e p e abili y o classes, pa simony
and en opy (Mu hén 2003; Nylund e al. 2007). The i indices a e p esen ed in de ail in
Table 1. The ou -class model had he lowes BIC alue and he lowes AIC alue. The e-
o e, we a o ed he ou -class model ha had he dis inc de elopmen pa e ns o all
he ou classes, highe en opy, and highe a e age pos e io p obabili y. The ou -class
model ga e us he la en classes ha ep esen low, dec easing, mode a e and high WLS,
espec i ely. The di e ences in basic and wo k- ela ed cha ac e is ics o he s udy sub-
jec s we e abula ed acco ding o ajec o ies o WLS wi h a signi icance le el o p<0.05.
The di e ences in con inuous a iables we e es ed using analysis o a iance and di -
e ences in ca ego ical a iables using he χ2 es . The mix u e modeling was done in
Mplus e sion 7.11 (Mu hen and Mu hen, 3463 S one A e., Los Angeles, CA) and o he
analyses we e done in STATA 14.0 (S a aCo p LP, College S a ion, Texas 77845, USA).
Table 1  Fi -indices  o   ajec o ies o  sa is ac ion in wo king li e among pos  o ice wo ke s
Classes BIC AIC En opy Pos e io p obabili y
2 27466.58 27371.44 0.73 0.92/0.93
3 26528.19 26382.68 0.73 0.91/0.89/0.86
4 25934.67 25738.79 0.75 0.87/0.88/0.92/0.84
BIC, Bayesian In o ma ion C i e ia; AIC, Akaike In o ma ion C i e ia
The quali a i e in e iew da a we e analyzed as na a i es o subjec i e WLS. In line wi h
Flanagan (1954), we ocused on c i ical inciden s, which s uc u e he na a i es by ei he
inc easing o educing he na a o ’s subjec i e WLS (see also Bu e ield e al. 2005).
C i ical inciden s a e common esou ces o na a i es. Because hey a e especially mean-
ing ul o na a o s, hey a e also well emembe ed. In c i ical inciden s echnique, key
inciden s a e iden i ied in a da a-d i en, bo om-up ashion om he ansc ibed in e iew
ma e ials (Bu e ield e al. 2005, p. 481–483). In his s udy, he analysis p oceeded as ol-
lows. The ansc ip ions and wo k sa is ac ion li eline d awings we e p ocessed oge he .
The c i ical inciden s desc ibed in he in e iews and ma ked on he d awings we e ca e-
go ized in o ones ha ook place when he na a o was 20–29, 30–39, 40–49 and 50–59
yea s old. Since he e alua ion o wo k sa is ac ion a he age o 60–69 was omi ed om
he quan i a i e s udy, his was also mi o ed in he quali a i e s udy. Nex , all inciden s
we e classi ied in o (a) ones ha inc eased WLS and (b) ones ha educed WLS.
Resul s
Fou dis inc pa hways o WLS we e iden i ied h ough quan i a i e analysis: high (27%),
mode a e (32%), dec easing (30%) and low (11%). The de elopmen al pa hways o WLS
o laid-o s a e shown in Figu e 1. WLS was mos ly inc easing o cons an om 15–29
yea s o 30–39 yea s o he low, mode a e and high ajec o y g oups. Howe e , o
he dec easing ajec o y g oup, WLS had al eady s a ed dec easing a 15–29 yea s and
sha ply dec eased a e 30–39 yea s o age. The membe ship o basic and wo k- ela ed
cha ac e is ics in ajec o ies o sa is ac ion in wo king li e a e desc ibed in Table 2.
O he o al 201 selec ed esponden s, 43% we e women and 57% we e men. The mean
26 E alua ion o Li elong Wo k Sa is ac ion Mii a Niska e al.
age o he esponden s was signi ican ly di e en among ajec o y g oups. Those in he
high and mode a e ajec o y g oup we e he oldes (60.2 ± 3.7 yea s, o high ajec o y
g oup) and he dec easing g oup was he younges (56.9 ± 3.4 yea s). The p opo ion o
whi e-colla wo ke s was 21% and he occupa ional class o he esponden s was no
signi ican ly di e en be ween WLS ajec o y g oups. The e was a signi ican di e ence
in pe cei ed heal h o he esponden s among he WLS g oups. A ound wo- i hs o
he esponden s had epo ed hei pe cei ed heal h o be poo and almos hal o hem
ep esen ed dec easing WLS. A ound 50% o he esponden s had basic schooling and
he e was no signi ican di e ence in educa ional s a us among he ajec o y g oups.
A ound 42% o he esponden s had 15–365 days o sickness absence and ou o ha
47% ep esen ed dec easing WLS.
Figu e 1 T ajec o ies o wo king li e sa is ac ion among pos se ice laid-o s in 2018.
Table 2 Dis ibu ion o basic and wo k- ela ed cha ac e is ics o he s udy popula ion in ou di -
e en ajec o ies o sa is ac ion in wo king li e (WLS) among Pos i G oup laid-o s
Cha ac e is ics o
he s udy popula ion
N
(201)
Membe ship o ajec o ies o sa is ac ion
in wo king li e
High
(n=54), %
Mode a e
(n=65), %
Dec easing
(n=61), %
Low
(n=21), %
p
- alueb
Agea
Gende
58.9±3.9 60.2±3.7 60.2±3.4 56.9±3.7 58.1±3.3 <0.001c
Women 85 31 34 28 7 0.434
Men 116 24 31 32 13
Occupa ional class
Whi e-colla 40 23 45 23 10 0.291
Blue-colla 153 28 29 33 11
No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020 27
Cha ac e is ics o
he s udy popula ion
N
(201)
Membe ship o ajec o ies o sa is ac ion
in wo king li e
High
(n=54), %
Mode a e
(n=65), %
Dec easing
(n=61), %
Low
(n=21), %
p
- alueb
Ch onic diseases
Yes 39 28 26 31 15 0.607
Pe cei ed Heal h
Good 56 23 45 20 10 0.005
Mode a e 67 34 35 21 10
Poo 77 22 22 47 9
Educa ional s a us
Academic deg ee 53 34 36 19 11 0.416
College T aining 57 23 33 37 7
Basic Schooling 91 25 30 33 12
Sickness absence
0–14 days 106 29 43 19 9 <0.001
15–365 days 77 22 19 47 12
Wo k hou s/weeka33.9±9.0 34.3±7.9 35.3±9.5 33.5±9.1 29.9±9.1 0.132c
Tenu e in yea sa28.2±12.1 30.1±11.3 27.9±12.3 27.8±12.2 25.5±12.9 0.498c
Equali y a job
No 72 22 24 43 11 0.016
Yes 125 29 39 22 10
Wo k abili y
Poo 79 25 24 43 8 <0.001
Mode a e 62 26 34 29 11
Good/Excellen 59 31 43 13 13
Job suppo
Low 122 25 27 41 9 <0.001
High 79 29 44 14 13
Job con ol
Low 126 21 25 44 10 <0.001
High 75 36 45 8 11
No e aexp essed as mean and SD (S anda d De ia ion); bde i ed om χ2- es ; cde i ed om analysis o
a iance
The a e age amoun o wo king hou s pe week among ou s udy subjec s was 33.9
hou s (SD 9.0) and he a e age numbe o enu e yea s in Pos i G oup was 28.2 yea s
(SD 12.1), howe e , none o hese we e signi ican ly di e en be ween WLS ajec o y
g oups. The p opo ion o good/excellen , mode a e and poo wo k abili y was 30%,
34 E alua ion o Li elong Wo k Sa is ac ion Mii a Niska e al.
he pa hways o WLS in a way p e ious s udies ha e no been able o do. Howe e , he
ac ha esponden s e alua ed hei gene al li e sa is ac ion jus be o e hey e alua ed
hei WLS engende s he possibili y o a ecall bias. None heless, he au ho s belie e
ha he use o mix u e modeling has con olled i o an ex en by mixing he analo-
gous esponses in o he sepa a e classes. In he quali a i e in e iew s udy, he use o
biog aphical li elines o map changes in WLS p oduced ich da a whe e in e iewees
e alua ed he inciden s ha inc eased o dec eased hei wo k sa is ac ion in de ail. This,
acco ding o Flanagan (1954), is c ucial o he c edibili y o he indings (also Bu e ield
e al. 2005). Rega ding he da a analysis, we we e in e es ed in in e iewees’ lay concep-
ions and how he esponden s hemsel es iden i ied he c i ical inciden s ha a ec ed
hei WLS. The eade is able o e alua e in elligibili y o he ca ego iza ion o hese
inciden s om he ex ac s p o ided in he esul s chap e .
Acknowledgemen s
This esea ch was suppo ed by he Kone Founda ion [5011cc] and The Finnish Wo k
En i onmen Fund [115440]. KC P was pa ly unded by a pe sonal wo king g an om
The Ella and Geo g Eh n oo h Founda ion g an , 2019. The au ho s a e g a e ul o Elisa
Vi kola (D.Soc.Sc) o he con ibu ion in da a gene a ion.
Re e ences
Allan, B. A., Dex e , C., Kinsey, R. & Pa ke , S. (2018). Meaning ul wo k and men al heal h:
job sa is aca ion as a mode a o , Jou nal o Men al Heal h 27(1): 38–44. doi: h ps://doi.
o g/10.1080/09638237.2016.1244718.
A ms ong-S assen, M. & Ca aneo, J. (2010). The e ec o downsizing on o ganiza ional
p ac ices a ge ing olde wo ke s, Jou nal o Managemen De elopmen 29(4): 344–363.
doi: h ps://doi.o g/10.1108/02621711011039150.
A e i, N., Abdullah, K. L., Wong, L. P. & Mazlom, R. (2014). Fac o s in luencing egis e ed
nu ses pe cep ion o hei o e all job sa is ac ion: a quali a i e s udy, In e na ional Nu s-
ing Re iew 61: 352–360. doi: h ps://doi.o g/10.1111/in .12112.
Bad an, M. & Ka a y, J. (2008). The e ec o job edesign on job sa is ac ion, esilience, com-
mi men and lexibili y: The case o an Egyp ian public sec o bank, In e na ional Jou nal
o Business Resea ch 8(3): 1–18.
Be glund, T., Seldén, D. & Halle öd, B. (2017). Fac o s A ec ing P olonged Wo king Li e o
he Olde Wo k o ce: he Swedish Case, No dic Jou nal o Wo king Li e S udies 7(1):
19–36. doi: h ps://doi.o g/10.18291/njwls. 7i1.81396.
Bojanowska, A. & Zalewska, A. (2016). Lay Unde s anding o Happiness and he Expe i-
ence o Well-being: A e some Concep ions o Happiness Mo e Bene icial han O he s?
Jou nal o Happiness S udies 17(2): 793–815. doi: h ps://doi.o g/10.1007/s10902-015-
9620-1.
Bussing, A., Bissels, T. & Fuchs, V. (1999). A Dynamic Model o Wo k Sa is ac ion: Qual-
i a i e App oaches, Human Rela ions 52 (8): 999–1028. doi: h ps://doi.o g/10.1023/
A:1016979523846.
Bu e ield, L. D., Bo gen, W. A., Amundson, N. E. & Maglio, A-S. T. (2005). Fi y yea s o he
c i ical inciden echnique: 1954–2004 and beyond, Quali a i e Resea ch 5(4): 475–497.
doi: h ps://doi.o g/10.1177/1468794105056924.

No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020 35
Chao, M. C., Jou, R. C., Liao, C. C. & Kuo, C. W. (2015). Wo kplace s ess, job sa is-
ac ion, job pe o mance, and u no e in en ion o heal h ca e wo ke s in u al Tai-
wan, Asia-Paci ic Jou nal o Public Heal h 27(2): NP1827–1836. doi: h ps://doi.o g/
10.1177/1010539513506604.
Da a, D. K., Gu h ie, J. P., Basuil, D. & Pandey, A. (2010). Causes and E ec s o Employ-
ee Downsizing: A Re iew and Syn hesis, Jou nal o Managemen 36(1): 281–348. doi:
h ps://doi.o g/10.1177/0149206309346735.
de Wesley, A. B. & Clemson, L. (1992). Job sa is ac ion issues: The ocus g oup app oach,
Aus alian Occupa ional The apy Jou nal 39: 7–15. doi: h ps://doi.o g/10.1111/j.
1440-1630.1992. b01762.x.
Dieke, A., Bende , C., Campbell, J. Cohen, R., e al. (2013). Main de elopmen s in he Pos al
Sec o (2010–2013), Wiik Consul , Final Repo .
Dolbie , C. L., Webs e , J. A., McCalis e , K. T., Mallon, M. W., S einha d , M. A. (2005).
Reliabili y and alidi y o a single-i em measu e o job sa is ac ion, Ame ican jou nal
o heal h p omo ion: AJHP 19(3): 194–198. doi: h ps://doi.o g/10.4278/0890-1171-
19.3.194.
Eläke u akeskus (2013). Suomen eläkejä jes elmän sopeu uminen eliniän pi enemiseen.
Eläkekysymys en asian un ija yö yhmän apo i, Helsinki: Eläke u akeskus.
Eu opean Commission (2010). P og ess and key challenges in he deli e y o adequa e and
sus ainable pensions in Eu ope, B ussels: Eu opean Commission.
Flanagan, J. C. (1954). The C i ical Inciden Technique, Psychological Bulle in 51: 327–358.
doi: h p://dx.doi.o g/10.1037/h0061470.
Fu unes, T. & Mykle un, R. (2010). Age disc imina ion in he wo kplace: Valida ion o he
No dic Age Disc imina ion Scale (NADS), Scandina ian Jou nal o Psychology 51(1):
23–30. doi: h ps://doi.o g/10.1111/j.1467-9450.2009.00738.x.
G un elde , J., Rispling, L. & No lén, G. (eds.) (2018a). S a e o he No dic Region 2018.
Theme 1: Demog aphy. Copenhagen: No dic Council o Minis e s.
G un elde , J., Rispling, L. & No lén, G. (eds.) (2018b). S a e o he No dic Region 2018.
Theme 2: Labou Fo ce. Copenhagen: No dic Council o Minis e s.
Ha konmäki, K., Ma ikainen, P., Lahelma, E., e al. (2009). In en ions o e i e, li e dissa is-
ac ion and he seubsequen isk o disabili y e i emen , Scandina ian Jou nal o Public
Heal h 37(3): 252–259. doi: h ps://doi.o g/10.1177/1403494808100273.
Hess, M. (2017). Rising P e e ed Re i emen Age in Eu ope: A e Eu ope’s Fu u e Pensione s
Adap ing o Pension Sys em Re o ms? Jou nal o Aging & Social Policy 29(3): 245–261.
doi: h ps://doi.o g/10.1080/08959420.2016.1255082.
I ella. Vuosike omus ja Y i ys as uu apo i (2011). (I ella annual epo and co po a e e-
sponsibili y epo 2011). h p:// uosike omus2011.pos i.com/Use Files/i ella/File/pd /
I ella_2011_Suomi.pd . (Accessed 3 Decembe 2018).
I ella. Vuosike omus ja Y i ys as uu apo i (2012). (I ella annual epo and co po-
a e esponsibili y epo 2012). h p:// uosike omus2012.i ella. i/Use Files/i ella/
File/2012/220838_I ella_ sk_2012_SUOMI.pd . (Accessed 3 Decembe 2018).
I ella. (2013). Vuosike omus. (I ella annual epo 2013). h p://annual epo 2013.i ella.
com/ ilebank/1332-I ella_Vuosike omus_2013.pd . (Accessed 3 Decembe 2018).
Jääskeläinen, A., Kaus o, J., Sei samo, J. e al. (2016). Wo kabili y index and pe cei ed wo k-
abili y as p edic o s o disabili y pension: a p ospec i e s udy among Finnish municipal
employees, Scandina ian Jou nal o Wo k, En i onmen & Heal h 42(6): 490–499. doi:
h ps://doi.o g/10.5271/sjweh.3598.
Kade o s, R., Schalle , J., Thång, P.-O. & Pes a , E. (2016). A i udes among Male and Fe-
male Uni e si y P o esso s, and o he Ca ego ies o Uni e si y Employees, o Wo king
up o and Beyond No mal Re i emen Age, No dic Jou nal o Wo king Li e S udies 6(1):
133–146. doi: h ps://doi.o g/10.19154/njwls. 6i1.4913.
36 E alua ion o Li elong Wo k Sa is ac ion Mii a Niska e al.
Kuoppala, J., Lamminpää, A., Väänänen-Tomppo, I. & Hinkka, K. (2011). Employee well-be-
ing and sick lea e, occupa ional acciden , and disabili y pension: a coho s udy o ci il
se an s, Jou nal o Occupa ional and En i onmen al Medicine 53(6): 633–640. doi:
h ps://doi.o g/10.1097/JOM.0b013e31821aa48c.
Lind o s, P. M., Me e oja, O. A., Töy y, S. M., Luukkonen, R. A., Elo ainio, M. J. & Leino, T.
J. (2007). Job sa is ac ion, wo k abili y and li e sa is ac ion among Finnish anes hesiolo-
gis s, Ac a Anaes hesiologica Scandina ica 51(7): 815–822. doi: h ps://doi.o g/10.1111/
j.1399-6576.2007.01343.x.
Locke, E. A. (1976). The na u e and causes o job sa is ac ion, In M.D. Dunne e (ed.), Hand-
book o indus ial and o ganiza ional psychology (pp. 1297–1351). New Yo k: Wiley.
Lunau, T., Bamb a, C., Eikemo, T. A., an de Wel, K. A. & D agano, N. (2014). A balancing
ac ? Wo k-li e balance, heal h and well-being in Eu opean wel a e s a es, The Eu opean
Jou nal o Public Heal h 24(3): 422–427. doi: h ps://doi.o g/10.1093/eu pub/cku010.
Luomanen, J., Lumme-Sand , K., Vi kola, E. & Nikande , P. (2018). Kadonnu yh eisö? Pos-
ilaisuus ikään y ien Pos in yön ekijöiden haas a elupuheessa, Työelämän u kimus –
A be sli s o skning 16 (3): 154–168.
Macky, K. A. (2004). O ganisa ional Downsizing and Redundancies: The New Zealand
Wo ke s Expe ience, New Zealand Jou nal o Employmen Rela ions 29(1): 63–87.
Mäkikangas, A. & Kinnunen, U. (2003). Psychosocial wo k s esso s and well-being: sel -
es eem and op imism as mode a o s in one-yea longi udinal sample, Pe sonali y and In-
di idual Di e ences 35: 537–557. doi: h ps://doi.o g/10.1016/S0191-8869(02)00217-9.
McMullin, J.A. & Ma shall, V.W. (2001). Ageism, age ela ions, and ga men indus y wo k in
Mon eal, Ge on ologis 41(1): 111–122. doi: h ps://doi.o g/10.1093/ge on /41.1.111.
McNeese-Smi h, D. K. (1999). A con en analysis o s a nu se desc ip ions o job sa is ac-
ion and dissa is ac ion, Jou nal o Ad anced Nu sing 29(6): 1332–1341. doi: h ps://doi.
o g/10.1046/j.1365-2648.1999.01018.x.
Meie , L. L., Tschudi, P., Meie , C. A., D o ak, C. & Zelle , A. (2015). When gene al p ac-
i ione s don’ eel app ecia ed by hei pa ien s: p ospec i e e ec s on well-being and
wo k- amily con lic in a Swiss longi udinal s udy, Family P ac ice 32(2): 181–186. doi:
h ps://doi.o g/10.1093/ amp a/cmu079.
Meye s, M. C., Adams, B. G., Sekaja, L., Buzea, C., Cazan, A-M., Go ea, M., e al. (2019).
Pe cei ed o ganiza ional suppo o he use o employees’ s eng hs and employee well-
being: a c oss-coun y compa ison, Jou nal o Happiness S udies 20(6): 1825–1841. doi:
h ps://doi.o g/10.1007/s10902-018-0026-8.
Moo e, K., C uickshank, M. & Haas, M. (2006). Job sa is ac ion in occupa ional he apy:
a quali a i e in es iga ion in u ban Aus alia, Aus alian Occupa ional The apy Jou nal
53: 18–26. doi: h ps://doi.o g/10.1111/j.1440-1630.2006.00539.x.
Mu hén, B. (2003). S a is ical and subs an i e checking in g ow h mix u e modeling: com-
men on Baue and Cu an, Psychological Me hods 8(3): 369–377. doi: h ps://doi.
o g/10.1037/1082-989X.8.3.369.
Na a o, M. & Sal e da, W. (2019). Ea ne posi ion and job and li e sa is ac ion: Do con-
ibu ions o he household income ha e he same e ec by gende and occupa ions.
Jou nal o Happiness S udies 20(7): 2227–2250. doi: h ps://doi.o g/10.1007/s10902-
018-0045-5.
Neale, B. (2019). Wha is Quali a i e Longi udinal Resea ch? London: Bloomsbu y.
Nikande , P. (2008). Wo king wi h T ansc ip s and T ansla ed Da a, Quali a i e Resea ch in
Psychology 5(3): 225–231. doi: h ps://doi.o g/10.1080/14780880802314346.
Nylund, K. L., Aspa ouho , T. & Mu hén, B. O. (2007). Deciding on he Numbe o Classes
in La en Class Analysis and G ow h Mix u e Modeling: A Mon e Ca lo Simula ion
S udy, S uc u al Equa ion Modeling: A mul idisciplina y Jou nal 14(4): 535–569. doi:
h ps://doi.o g/10.1080/10705510701575396.
No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020 37
OECD (2017). OECD ecommends Finland o do mo e o help olde people s ay in wo k.
h p://www.oecd.o g/ inland/oecd ecommends inland odomo e ohelpolde peoples ayin-
wo k.h m. (Accessed 4 Oc obe 2017).
Pagán, R. (2013). Job Sa is ac ion and Domains o Job Sa is ac ion o Olde Wo ke s wi h
Disabili ies in Eu ope, Jou nal o Happiness S udies 14(3): 861–891. doi: h ps://doi.
o g/10.1007/s10902-012-9359-x.
Pie iäinen, J-P. (1998). Pos ia kaikille. Suomen Pos in a ina 1638–1998. Helsinki: Edi a.
Pos i (2014). Vuosike omus 2014. (Pos i G oup annual epo 2014). h p://annual epo 2014.
pos i.com/ ilebank/1447-Pos i_Vuosike omus_2014.pd . (Accessed 3 Decembe 2018).
Pos i (2015). Vuosike omus 2015. (Pos i G oup annual epo 2015). h p://annual epo 2015.
pos i.com/ ilebank/1230-POSTI_Vuosike omus_2015.pd . (Accessed 3 Decembe 2018).
Pos i (2016). Vuosike omus 2016. (Pos i G oup annual epo 2016). h p://annual e-
po 2016.pos i.com/ ilebank/800-Pos i_VSK16_FI.pd . (Accessed 3 Decembe 2018).
Qian, Y. & Fan, W. (2019). Men and women a wo k: Occupa ional gende composi ion and
a ec i e well-being in he Uni ed S a es. Jou nal o Happiness S udies 20(7): 2077–2099.
doi: h ps://doi.o g/10.1007/s10902-018-0039-3.
Ramoo, V., Abdullah, K. L., Piaw, C. Y. (2013). The ela ionship be ween job sa is ac ion and
in en ion o lea e cu en employmen among egis e ed nu ses in a eaching hospi al.
Jou nal o Clinical Nu sing 22: 3341–3152. doi: h ps://doi.o eg/10.1111/jocn.12260.
Sagie, A. (1998). Employee Absen eeism, O ganiza ional Commi men , and Job Sa is ac ion:
Ano he Look, Jou nal o Voca ional Beha io 52: 156–171. doi: h ps://doi.o g/10.1006/
j be.1997.1581.
Sieg is , J., Wah endo , M., Von dem Knesebecko, O., Jü ges, H. & Bö sch-Supan, A. (2007).
Quali y o wo k, well-being, and in ended ea ly e i emen o olde employees: Baseline
esul s om he SHARE s udy, Eu opean Jou nal o Public Heal h 17: 62–68. doi: h ps://
doi.o g/10.1093/eu pub/ckl084.
Shimazu, A. & Schau eli, W. B. (2009). Is wo kaholism good o bad o employee well-being?
The dis inc i eness o wo kaholism and wo k engagemen among Japanese employees,
Indus ial Heal h 47: 495–502. h ps://doi.o g/10.2486/indheal h.47.495.
Solem, P-E. (2016). Ageism and age disc imina ion in wo king li e, No dic Psychology 68(3):
160 –175. doi: h ps://doi.o g/10.1080/19012276.2015.1095650.
Spec o , P. E. (1997). Job sa is ac ion: Applica ion, assessmen , causes and consequences.
Thousand Oaks: Sage.
S ynen, D., Jansen, N. W., Slangen, J. J. & Kan , I. (2016). Impac o de elopmen and accom-
moda ion p ac ices on olde wo ke s’ job cha ac e is ics, p olonged a igue, wo k engage-
men , and e i emen in en ions o e ime, Jou nal o Occupa ional and En i onmen al
Medicine 58(11): 1055–1065. doi: h ps://doi.o g/10.1097/JOM.0000000000000853.
Tay, L. & Ha e , J. K. (2013). Economic and labo ma ke o ces ma e o wo ke
well-being, Applied Psychology: Heal h & Well-Being 5(2): 193–308. doi: h ps://doi.
o g/10.1111/aphw.12004.
an den Be g, T. I. J., Elde s, L. A. M. & Bu do , A. (2010). In luence o Heal h and Wo k on
Ea ly Re i emen , Jou nal o Occupa ional and En i onmen al Medicine 52(6): 576–583.
doi: h ps://doi.o g/10.1097/JOM.0b013e3181de8133.
Vi anen, M., Ki imäki, M., Elo ainio, M., Vah e a, J. & Fe ie, J. E. (2003). F om insecu e
o secu e employmen : changes in wo k, heal h, heal h ela ed beha iou s, and sickness
absence, Occupa ional and En i onmen al Medicine 60(12): 948–953. doi: h ps://doi.
o g/10.1136/oem.60.12.948.
on Bonsdo , M. B., Sei samo, J., on Bonsdo , M. E., e al. (2012). Job s ain among
blue-colla and whi e-colla employees as a de e minan o o al mo ali y: a 28-yea
popula ion based ollow-up, BMJ Open 2(2). doi: h p://dx.doi.o g/10.1136/bmjop-
en-2012-000860.
38 E alua ion o Li elong Wo k Sa is ac ion Mii a Niska e al.
Waddimba, A. C., Sc ibani, M., K upa, N., May, J. J. & Jenkins, P. (2016). F equency o sa is-
ac ion and dissa is ac ion wi h p ac ice among u al-based, g oup employed physicians
and non-physician p ac i ione s, BMC Heal h Se ices Resea ch 16(1): 613. doi: h ps://
doi.o g/10.1186/s12913-016-1777-8.
Wise, D. (2010). Facili a ing longe wo king li es: in e na ional e idence on why and how,
Demog aphy 47 (supplemen ): S131–S149. doi: h ps://doi.o g/10.1353/dem.2010.0000.
Zembylas, M. & Papanas asiou, E. (2006). Sou ces o eache job sa is ac ion and dissa is ac-
ion in Cyp us, Compa e: A Jou nal o Compa a i e and In e na ional Educa ion 36(2):
229–247. doi: h ps://doi.o g/10.1080/03057920600741289.
No dic jou nal o wo king li e s udies Volume 10 ❚ Numbe 1 ❚ Ma ch 2020 39
Appendix
Table A1 Quali a i e da a used in he p ojec
Quali a i e da a Amoun
Face- o- ace in e iews 40 (1s ound) + 32
(2nd ound)
Telephone in e iews 107
Video dia ies 216
Jou nals 17
e-mails 18
Figu e A1 Coo dina e plane gi en o he in e iewees