Wo schack, Philip; Sam leben, Clai e
A icle — Published Ve sion
C osse s in a Segmen ed Labou Ma ke : Occupa ional
Ad ancemen and Wage Changes om Semi-Skilled and
Unskilled Jobs
Wo k, Employmen and Socie y Wo k, Employmen and Socie y
P o ided in Coope a ion wi h:
WZB Be lin Social Science Cen e
Sugges ed Ci a ion: Wo schack, Philip; Sam leben, Clai e (2024) : C osse s in a Segmen ed Labou
Ma ke : Occupa ional Ad ancemen and Wage Changes om Semi-Skilled and Unskilled Jobs, Wo k,
Employmen and Socie y Wo k, Employmen and Socie y, ISSN 1469-8722, Sage, Thousand Oaks,
CA, Vol. 39, Iss. 2, pp. 496-515,
h ps://doi.o g/10.1177/09500170241275861
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/310962.2
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
h ps://doi.o g/10.1177/09500170241275861
Wo k, Employmen and Socie y
2025, Vol. 39(2) 496 –515
© The Au ho (s) 2024
A icle euse guidelines:
sagepub.com/jou nals-pe missions
DOI: 10.1177/09500170241275861
jou nals.sagepub.com/home/wes
C osse s in a Segmen ed
Labou Ma ke : Occupa ional
Ad ancemen and Wage
Changes om Semi-Skilled and
Unskilled Jobs
Philip Wo schack
WZB Be lin Social Science Cen e & Weizenbaum Ins i u e, Be lin
Clai e Sam leben
P ognos, Ge many
Abs ac
How he upwa d mobili y chances o wo ke s in unskilled o semi-skilled jobs a e shaped by
in luences a he o ganisa ional and sec o al le el emains an open ques ion. This a icle aims
o close his esea ch gap by examining he ole o in e nal labou ma ke cha ac e is ics in he
p omo ion p ospec s and wage inc eases o wo ke s in semi-skilled and unskilled posi ions. The
hypo heses a e de i ed om dual and segmen ed labou ma ke heo y. Reg ession analyses
based on linked-employe -employee-da a (LIAB), co e ing 44,024 wo ke s in semi-skilled and
unskilled posi ions om 2005 o 2010, unde line he impo ance o he in e nal labou ma ke .
A conside able sha e o wo ke s mo ed o skilled posi ions h ough company change. Fo he
wo ke s who s ayed wi h he company, ca ee ad ancemen s we e associa ed wi h egula
aining in es men s and o malised egula ions a he company le el. Collec i e ag eemen s, in
con as , we e associa ed wi h lowe chances o upwa d mobili y, bu highe wages.
Keywo ds
(in)equali y and disc imina ion, policy and egula ion, skills and aining
In oduc ion
In Eu opean poli ical and scien i ic deba es, in-company aining is o en seen as an
ins umen o achie e bo h p oduc i i y gains o he company and imp o emen s in
employmen and p omo ion oppo uni ies o employees, especially o disad an aged
Co esponding au ho :
Philip Wo schack, WZB Be lin Social Science Cen e , Reichpie schu e 50, Be lin, 10785, Ge many.
Email: [email p o ec ed]
1275861WES0010.1177/09500170241275861Wo k, Employmen and Socie yWo schack and Sam leben
esea ch-a icle2024
A icle
Wo schack and Sam leben 497
labou ma ke g oups (A ulampalm e al., 2004; BMBF, 2015: 46; Bogai e al., 2017;
Moh e al., 2016; OECD, 2004; Wo schack, 2020a, 2020b).
Howe e , a e iew o he exis ing esea ch on con inuing oca ional aining shows
ha he ac ual e ec s o in-company aining on income, employabili y and upwa d
mobili y a e a om clea (Asplund, 2005: 69). Al hough he e ha e been se e al con-
sis en esea ch indings ha p o e posi i e e ec s om in-company aining on wage
inc eases and employabili y (Ramos and Ha is, 2012), he o e all numbe o empi ical
s udies is a he small. Many s udies ha e also been limi ed o speci ic coun ies, sec o s
o companies, and esul s ha e been inconsis en (Becke and Hecken, 2009: 378;
Ehle , 2017; Hansson, 2008: 22). Rega ding heo ies in he ield, he ole o labou
ma ke ins i u ions and collec i e ac o s is o en neglec ed (G imshaw e al., 2017;
K ings, 2020). This a icle aims o ill his gap in he esea ch. I is guided by he idea
ha labou ma ke ad ancemen s o wo ke s in semi-skilled o unskilled jobs a e no
only shaped by indi idual aining beha iou , bu also by he ins i u ional s uc u es o
he in e nal labou ma ke . The c ucial ques ion is in which way and o wha ex en a e
hese wo ke s’ upwa d mobili y in luenced by ins i u ional cha ac e is ics a he com-
pany and sec o al le el?
One p ominen app oach add essing he o ganisa ional and ins i u ional con ex is he
dual o segmen ed labou ma ke heo y (Dü sch and S uck, 2014; Emmenegge e al.,
2012; Köhle e al., 2008; Sengenbe ge , 1982; Seo, 2021). I assumes ha he labou
ma ke is s uc u ed in o di e en sec o s o segmen s: a p ima y sec o cha ac e ised by
jobs wi h good condi ions ega ding employmen secu i y, pay, aining and p omo ion
p ospec s, and a seconda y sec o wi h a he bad jobs, cha ac e ised by low skill
demands, sho - e m employmen , limi ed upwa d mobili y chances and lowe wages.
The p ima y segmen is ypically p esen in la ge companies wi h a mo e egula ed
in e nal labou ma ke and hie a chical p omo ion and ewa d sys ems. In his ypology,
he ques ion o he (im)pe meabili y o he di e en segmen s emains unclea . Acco ding
o he heo e ical app oach, wo ke s’ abili ies o mo e om he second o he i s seg-
men a e s ongly es ic ed by he segmen ed s uc u e o he labou ma ke .
A ecen s udy in Ge many has p o ided e idence o upwa d mobili y o wo ke s in
he low-skilled segmen , indica ing a leas some deg ee o pe meabili y. Acco ding o
his s udy, abou hal o he people wi hou oca ional quali ica ions we e employed as
skilled wo ke s (Bü mann and Wiek, 2018). This aises he ques ion o how hey could
ge in o skilled posi ions despi e he dualised labou ma ke s uc u e and Ge man oca-
ional aining sys em, which a aches g ea impo ance o o mal quali ica ions. Two
possible pa hways – he subjec o his a icle – a e ansi ions in he ex e nal and in e nal
labou ma ke .
This a icle aims o explo e hese pa hways u he . Fi s , he ole o employe change
is in es iga ed. How does i ela e o he upwa d mobili y chances and ea nings o wo k-
e s in semi-skilled o unskilled jobs? Second, he impac o he ins i u ional shape o he
in e nal labou ma ke is explo ed in e ms o o malised HR policies, egula aining
in es men s, s uc u es o employee ep esen a ion and collec i e ag eemen s.
Theo e ically, his s udy builds on a c i ical ecep ion o dual labou ma ke heo y
ha acknowledges he impo ance o legal egula ions, collec i e ac o s, o ganisa ional
policies and powe ela ions (Emmenegge e al., 2012; G imshaw e al., 2017; K ings,
498 Wo k, Employmen and Socie y 39(2)
2020) while also indica ing mo e complex pa e ns o segmen a ion (Lukac e al., 2019;
Seo, 2021).
Rega ding he empi ical base o his s udy, linked-employe -employee-da a (LIAB)
om he Ge man Ins i u e o Employmen Resea ch (IAB) was analysed (see Ellgu h
e al., 2014). The sample comp ises o 44,024 ull- ime wo ke s in semi-skilled and
unskilled posi ions om 1272 companies; 26,853 o hem did no change hei employe
du ing he i e-yea obse a ion pe iod (2005–2010).
One las ema k on how he e m ‘skills’ is unde s ood in his s udy. Voca ional skills
a e no a neu al concep ; hey a e a socially cons uc ed concep (Wa hu s e al., 2017).
In his a icle, we use he e m in he (na owe ) sense o nominal skills (Rigby and
Sanchis, 2021: 23), de ined as he skills employe s a e p epa ed o ecognise and ewa d
(in con as o he en i e y o wo ke s’ knowledge, skills and expe ience). Semi-skilled
and unskilled jobs indica e bundles o wo k asks ha a e ela i ely easy o access, lea n
and pe o m wi hou longe pe iods o oca ional aining.
In luences o upwa d mobili y in low-skilled jobs: S a e o
esea ch
Fo mal aining ha leads o ecognised oca ional quali ica ions plays a c ucial ole
in he highly ins i u ionalised Ge man oca ional educa ion sys em (Thelen, 2014).
Though he idea o li elong lea ning has been s ongly emphasised by he public and
scien i ic deba e as he answe o an expanding ange o p ac ical and policy chal-
lenges, wo ke s wi hou adequa e ini ial oca ional quali ica ions ge ew ‘second
chances’ in he Ge man oca ional educa ion sys em o achie e o mal p o essional
quali ica ions du ing hei employmen ca ee s (BMAS and BMBF, 2019; Rahne ,
2014). Howe e , wi h he Quali ica ion Oppo uni y Ac (since 2019) he Ge man go -
e nmen has p o ided new inancial suppo ega ding o mal aining measu es o
his g oup.
Unlike o mal aining p og ammes – which a e ocused on gene al skills and lead o
o icial p o essional quali ica ions – in-company aining measu es include in o mal
aining op ions such as cou ses, ins uc ion o wo kplace-based o ms o lea ning. These
a e ini ia ed by he company and ake place in he company con ex (Beich e al., 2004).
Making up a ound 65% o con inuing oca ional aining p og ammes, hey ep esen
he la ges segmen o con inuing aining (BMBF, 2021). Howe e , hey ocus on com-
pany-speci ic skill equi emen s and adap a ion measu es ha a e ela i ely sho e m
and do no usually lead o ecognised oca ional quali ica ions. Long- e m aining p o-
g ammes ha o e employees oppo uni ies o oca ionally eo ien o ca ch up on oca-
ional quali ica ions a e a ely ep esen ed (BMBF, 2015). Consequen ly, wo ke s’
chances o achie e p o essional quali ica ions ia con inuing aining a e gene ally lim-
i ed, pa icula ly when hey a e wo king in jobs wi h low skill demands. Many o hese
wo ke s epo subs an ial conce ns ega ding whe he con inuing oca ional aining
will ac ually pay o and lead o g ea e income, job secu i y o p omo ion oppo uni ies
(Osiande and S ephan, 2018).
Wo schack and Sam leben 499
A e iew o he exis ing esea ch on con inuing oca ional aining con i ms ha he
ac ual e ec s o in-company aining on income, employabili y and upwa d mobili y a e
a om clea (Asplund, 2005: 69). Al hough he e a e se e al consis en esea ch ind-
ings ha p o e posi i e e ec s om in-company aining on wage inc eases and employ-
abili y (Ramos and Ha is, 2012), he o e all numbe o empi ical s udies is a he small.
Many s udies a e limi ed o speci ic coun ies, sec o s o companies, and hei esul s a e
some imes inconsis en (Becke and Hecken, 2009: 378; Hansson, 2008: 22). Mo eo e ,
many s udies do no accoun o possible ade-o s be ween he di e en e u ns om
aining (Hansson, 2008: 22; OECD, 2004) such as income, upwa d occupa ional mobil-
i y o employmen secu i y.
Va ious s udies ha e examined he ole o con inuing oca ional aining o ca ee
ad ancemen oppo uni ies in gene al, showing posi i e impac s on ca ee ad ancemen s
o employmen secu i y (Ebene and Ehle , 2018; G een e al., 2000; Hansson, 2008;
Pannenbe g, 1995). The ques ion o how on- he-job aining a ec s he occupa ional
mobili y o wo ke s in semi-skilled and unskilled jobs has only been add essed by a ew
s udies. E e sson’s (2004) indings show ha employees wi h low quali ica ions a e less
likely o pa icipa e in aining p og ammes ha a e ele an o p omo ions wi hin he
company. Sande s and de G ip (2004) examined he in e play be ween pa icipa ion in
aining, job ask p o ile and labou ma ke mobili y o low-skilled wo ke s. They con-
cluded ha , o his g oup, con inuing oca ional aining is only bene icial on he in e -
nal labou ma ke .
A numbe o s udies ha e demons a ed he posi i e e ec s o in-company aining on
income de elopmen (Becke and Schömann, 1996; Ehle , 2017; OECD, 2004; Pa en ,
1999; Ramos and Ha is, 2012; Zwick, 2003). Some epo ed a ying e u ns om ain-
ing depending on he size o he company (Ehle , 2017; Pannenbe g, 2008), occupa-
ional s a us o he wo ke s (E e sson, 2004), unding sou ce (Boo h and B yan, 2002;
Hansson, 2008), company change (Becke and Schömann, 1996; Boo h and B yan,
2002; OECD, 2004), ype o aining (manda o y o no ) and he labou ma ke segmen
(in e nal o no ) (Ehle , 2017).
Theo y and hypo heses: Upwa d mobili y in segmen ed
labou ma ke s
In he li e a u e on con inuing aining, heo e ical app oaches ha ocus on human
capi al (Becke , 1975) domina e, emphasising he impo ance o indi idual cha ac e -
is ics and he labou supply side. The ole o ex e nal and in e nal labou ma ke cha -
ac e is ics in he ca ee ad ancemen s o wo ke s in semi-skilled o unskilled jobs has
been widely neglec ed.
One impo an app oach emphasising he o ganisa ional and ins i u ional con ex is
he dual o segmen ed labou ma ke heo y (Dü sch and S uck, 2014; Emmenegge
e al., 2012; Köhle e al., 2008; Sengenbe ge , 1982). A he co e o his heo y is he
assump ion ha jobs clus e in e ms o (bad) pay, (lack o ) employmen s abili y and
(limi ed) mobili y oppo uni ies, leading o a dualised labou ma ke s uc u e
(Emmenegge e al., 2012).
500 Wo k, Employmen and Socie y 39(2)
O he schola s ha e ques ioned his assump ion. The dis inc ion be ween p ima y and
seconda y sec o s migh be oo simplis ic, as labou ma ke segmen s ha e become mo e
complex (K ings, 2020: 529). Mo eo e , he dis inc ion o he di e en segmen s is usu-
ally educed o employmen con ac s, o e looking how hey combine wi h o he aspec s
o p eca iousness like income, job p ospec s and pe cei ed insecu i y (Lukac e al.,
2019; Seo, 2021). A ecen s udy ound e idence ha a conside able sha e o wo ke s is
cha ac e ised by low income and lack o job p ospec s, despi e ha ing s able employ-
men s a us (Seo, 2021: 494). E en ually, he ole o ins i u ions and ac o s in shaping
di e en labou ma ke segmen s is neglec ed (K ings, 2020: 530), unde lining he need
o explo e he impac o (de-) egula ion policies, collec i e ag eemen s and unions’ ba -
gaining powe (Appelbaum e al., 2010; P osse , 2016).
This s udy cen es on Ge many. Ge many has been cha ac e ised as a p ominen case
o a coo dina ed ma ke economy wi h a s ong dual app en iceship sys em, cen alised
collec i e ba gaining and a s ong o ganisa ion o con inuing aining ac i i ies a he
company le el (Allaa e al., 2009: 105; Es é ez-Abe, 2005). Howe e , co e age in col-
lec i e ba gaining has been sh inking since he 1990s due o inc easing d opou s o com-
panies and inc easing numbe s o p eca ious wo ke s, pa icula ly in he se ice sec o
(Swank e al., 2008). Consequen ly, he e is e idence o a g owing di ide be ween he
co e age o wo ke s who do skilled wo k in he manu ac u ing sec o and he la ge num-
be s o wo ke s in semi-skilled and unskilled jobs who a e no co e ed by he collec i e
ba gaining sys em (Doellgas e al., 2018; Thelen, 2009: 482; Thelen, 2014).
Likewise, he dual app en iceship sys em has become mo e segmen ed. The g owing
se ice economy con o ms less o en o he s anda ds o he adi ional dual aining
sys em. The e is e idence ha he ocus on s anda dised po able skills and exchange-
able quali ica ions in occupa ional labou ma ke s has been pa ly eplaced by a sys em
mo e gea ed owa ds company-speci ic skill needs (Thelen, 2009), less demanding
( wo-yea ) app en iceships mainly in he se ice economy o school-based oca ional
acks (as in he heal h and ca e sec o ) (Es é ez-Abe, 2005; Thelen and Busemeye ,
2008). The adi ional dual app en iceship sys em is cha ac e ised by la ge up on
in es men s in gene al and company-speci ic skills and s ic , o malised skill equi e-
men s ega ding p omo ion. Fo wo ke s in semi-skilled and unskilled jobs, con inuous
non- o mal aining in es men s o e he li e cou se and mo e pe meable employmen
ajec o ies gain g ea e impo ance (Thelen, 2014) and migh also os e g ea e p omo-
ion chances.
In e nal labou ma ke cha ac e is ics and upwa d mobili y chances
These de elopmen s show how he quali y o wo k, employmen and mobili y chances
in di e en labou ma ke segmen s is shaped by o ganisa ional policies, legal and sec o-
al egula ions and powe ela ions (G imshaw e al., 2017; K ings, 2020). When we
apply his idea o he ques ion o how wo ke s can ad ance om semi-skilled o unskilled
jobs o skilled posi ions, he ocus shi s om indi idual aining beha iou o he cha -
ac e is ics o he ex e nal and in e nal labou ma ke . How does he p esence o absence
o inclusi e egula ions and collec i e in e es ep esen a ion a he company, sec o al
and na ional le el shape he upwa d mobili y chances o his g oup o wo ke s?
Wo schack and Sam leben 501
We expec ha wo ke s’ chances o mo ing om semi-skilled o unskilled posi ions
o highe (skilled) unc ions will depend on he ins i u ional s uc u es o he company’s
in e nal labou ma ke . S aying wi h one employe does no necessa ily help in mo ing
up in he in e nal labou ma ke ; e en less when wo ke s a e in a weak posi ion – as is
he case o wo ke s in unskilled o semi-skilled jobs. We expec ha he shape and s uc-
u e o he in e nal labou ma ke is an impo an de e minan he e. I i does no p o ide
he esou ces and ins i u ional s uc u es o suppo upwa d mobili y o hose wo king
in he low-skilled job segmen , a change o employe migh be he be e al e na i e o
imp o e hei labou ma ke chances. This basic expec a ion is b oken down and es ed
in he ollowing h ee hypo heses.
Fi s , we expec ha a conside able sha e o wo ke s in semi-skilled o unskilled posi-
ions will no ad ance o skilled posi ions by s aying in he company (hypo hesis H1).
We assume ha his will ypically be he case when i is mo e di icul o mo e up in he
in e nal labou ma ke due o a s ong deg ee o in e nal segmen a ion.
Second, we expec ha he upwa d mobili y chances o wo ke s in semi-skilled o
unskilled posi ions will a y wi h he ins i u ional cha ac e is ics o he in e nal labou
ma ke ma e (hypo hesis H2).
Ins i u ional labou ma ke in luences a he company and
sec o al le el
Following he c i ical e iew o dual and segmen ed labou ma ke heo y (Dü sch and
S uck, 2014; Köhle e al., 2008; K ings, 2020; Sengenbe ge , 1982; Seo, 2021), we
expec ha he p omo ion chances o wo ke s in semi-skilled o unskilled jobs a e in lu-
enced by companies’ aining policies, he size o he in e nal labou ma ke and i s egu-
la ion ia company ag eemen s and in e es ep esen a ion s uc u es. La ge companies
(H2.1), companies wi h egula o high aining in es men s (H2.2), companies wi h
o malised HR policies (H2.3) and companies wi h in e es ep esen a ion s uc u es
(H2.4) a e associa ed wi h highe chances o ca ee ad ancemen o wo ke s in low-
skilled jobs.
The scope o p o essional o inancial ad ancemen is usually g ea e in la ge com-
panies han in small companies due o he mo e ex ensi e esou ces and la ge numbe
o a ailable posi ions. This should be bene icial o wo ke s in semi-skilled and unskilled
jobs in e ms o hei oppo uni ies o ca ee ad ancemen .
As sugges ed by many policy make s and schola s who emphasise he impo ance o
con inuing aining o upwa d mobili y (o en e e ing o human capi al heo y), egu-
la and/o high in-company aining in es men s o wo ke s in semi-skilled o unskilled
jobs should inc ease hei chances o pe o ming skilled asks and being p omo ed
acco dingly.
W i en egula ions and p ocedu es in HR wo k – in he o m o a ge ag eemen s,
pe o mance app aisals, job desc ip ions, p ocedu es o illing posi ions and plans o
s a ing equi emen s – inc ease anspa ency and he amoun o in o ma ion on he
ac ual pe o mance o indi idual employees, u u e labou and skill demands and a -
ge ed pe sonnel de elopmen . Wo ke s in semi-skilled and unskilled jobs a e likely o
502 Wo k, Employmen and Socie y 39(2)
bene i om his in e ms o a ge ed pe sonal de elopmen and wo kplace lea ning o
ill acan , skilled posi ions.
In e es ep esen a ion s uc u es play an impo an ole in ebalancing he powe
asymme y be ween managemen and wo k o ce in a ou o employees and o e com-
ing ba ie s o upskilling. In e es ep esen a ion bodies a e asked wi h ep esen ing he
in e es s o all g oups o employees, including wo ke s in semi-skilled and unskilled
jobs. This includes hei access o company ewa ds and p omo ions. Semi-skilled and
unskilled wo ke s in companies should he e o e be mo e likely o ealise occupa ional
s a us imp o emen s i in e es ep esen a ion s uc u es a e in place.
A he sec o al le el, wo impo an ins i u ional con ex cha ac e is ics a e o impo -
ance (hypo hesis H3): collec i e ba gaining co e age and he di e ences be ween
‘co e’ e sus ‘pe iphe al’ sec o s o he Ge man sys em o indus ial ela ions – hese
go along wi h s ong e sus weak s anda disa ion and o malisa ion o quali ica ions
and ca ee s.
Collec i e ba gaining egula ions se uni o m s anda ds o adequa e sala ies depend-
ing on asks o quali ica ions. Wages a e dis ibu ed acco ding o collec i ely ag eed
s anda ds depending on he equi ed skills (pay g oup) and ac i i y pe o med (pay
g ade). In his espec , an inclusi e wage-se ing e ec o collec i e ag eemen s would
be expec ed (Eichho s and Ma x, 2015: 45), ensu ing he pa icipa ion o wo ke s in
semi-skilled o unskilled jobs (H3.1). On he o he hand, due o s onge collec i ely
ba gained egula ions, o mal oca ional quali ica ions a e likely o play a g ea e ole in
in e nal p omo ion sys ems, which should make upwa d mobili y wi hin he amewo k
o non- o mal aining ac i i ies mo e di icul (H3.2). In his con ex , endencies owa ds
closu es on he pa o ade unions may also play a ole, inso a as hese en ail a emp s
o secu e quali y and income s anda ds o ce ain unc ions ia o malised quali ica ion
equi emen s and egula e access o hese posi ions.
Rega ding sec o al di e ences, we expec ha in he ‘co e’ sec o o he Ge man sys-
em o indus ial ela ions wi h a he s ong unions (manu ac u ing), he e will be be e
wage de elopmen o wo ke s in semi-skilled o unskilled posi ions due o a mo e inclu-
si e wage-se ing con ex (H3.3). In con as , he ansi ion om unskilled o semi-
skilled jobs o skilled jobs will be mo e es ic ed due o s onge s anda disa ion and
o malisa ion o quali ica ions and ca ee pa hways in his highly egula ed sec o com-
pa ed wi h o he sec o s (pa icula ly he se ice economy) (H3.4).
Resea ch design
Da a
The hypo heses we e es ed using LIAB da a om he IAB (Ellgu h e al., 2014; Heining
e al., 2012). Da a we e accessed ia a gues isi o he Resea ch Da a Cen e o he
Ge man Fede al Employmen Agency (BA) a he Ins i u e o Employmen Resea ch
(FDZ) as well as con olled emo e da a p ocessing. The LIAB da a (LIAB longi udinal
model 9310) combines ich company in o ma ion om he annual su ey wa es o he
IAB company panel wi h egis e da a om he BA.
Wo schack and Sam leben 503
Ou sample o he LIAB da a (2005–2010) includes all employees who we e employed
as ‘semi-skilled and unskilled wo ke s’ a he beginning o he s udy pe iod (2005). The
decisi e ac o is he e o e he occupa ional s a us in ela ion o he ac i i y pe o med
(and no he o mal occupa ional quali ica ion). Un o una ely, mo e ecen wa es o he
LIAB could no be used o his s udy as he cen al occupa ional s a us cha ac e is ic o
in e es (‘semi-skilled and unskilled wo ke s’) was no longe eco ded.
The sample o wo ke s in semi-skilled and unskilled jobs included in he s udy a e all
in ull- ime employmen . The sample is he e o e no ep esen a i e o all employees in
Ge many. Al hough a ixed-e ec s model would ha e been possible (as con i med by
es ing pooled o dina y leas squa es [OLS] eg ession agains leas squa es dummy a i-
ables [LSDV]), we decided on he pooled OLS eg ession model o analy ical easons
and da a- ela ed limi a ions. The ocus o ou s udy is on mo e ime-cons an explana o y
a iables – such as s uc u al and ins i u ional company cha ac e is ics – ha do no a y
much o e he obse ed pe iod o i e yea s and would ha e been excluded in ixed-
e ec s models. Ins i u ional in luences equi e ime o un old hei e ec s; hese could
no be obse ed o a longe pe iod due o da a limi a ions.
A longe obse a ion pe iod was no possible as some o ou key a iables (mos
impo an ly occupa ional s a us o iden i y wo ke s in skilled o semi-skilled jobs) we e
no used in la e wa es o he panel. To use he longi udinal s uc u e o he da a, we
he e o e decided o use pooled OLS and logis ic eg essions o e a i e-yea obse a-
ion pe iod. Al hough we canno make any s a emen s abou causal e ec s, ou s udy
aims o iden i y co ela ions be ween o ganisa ional cha ac e is ics (p ima ily he ins i-
u ional con ex ) and he upwa d mobili y chances o wo ke s in unskilled o semi-skilled
jobs. A e checking o in aclass co ela ions, which showed mino di e ences ega d-
ing ou coe icien s, we conside ed a simple model wi h clus e ed s anda d e o s as
su icien .
Me hod
By using eg ession analyses, possible p omo ions o semi-skilled and unskilled wo ke s
in he in e nal labou ma ke we e measu ed by occupa ional ad ancemen om 2005 o
2010 om semi-skilled/unskilled wo ke o skilled wo ke , mas e c a spe son, o e-
man/woman o whi e-colla wo ke . The income de elopmen in he pe iod unde s udy
(measu ed by he di e ence in ‘daily pay’ be ween 2005 and 2010) and he employmen
s a us (employmen subjec o social secu i y con ibu ions) a he end o he s udy pe iod
we e also included as addi ional indica o s.
Va iables
We measu ed egula in-company aining in es men s o he g oup o wo ke s in semi-
skilled and unskilled jobs, in es iga ing whe he he company p o ided suppo o con-
inuing aining o (a leas one) semi-skilled and unskilled wo ke du ing he pe iod
unde s udy, be i inancial suppo o ime o . Since he companies’ con inuing aining
ac i i ies we e measu ed a h ee poin s in ime (2005, 2007, 2010), he a iable can
assume alues be ween 0 (no in es men in con inuing aining o semi-skilled and
510 Wo k, Employmen and Socie y 39(2)
g oup o semi-skilled and unskilled wo ke s, who a e o en in mo e p eca ious
employmen .
Ano he p oblem wi h he da a is ha he in o ma ion on occupa ional s a us is
employe da a o which he accu acy o imeliness o he in o ma ion was no e i ied.
Consequen ly, we can assume ha he da a did no accoun o all changes in occupa-
ional s a us and ha occupa ional s a us imp o emen s a e he e o e unde es ima ed in
he analyses. Howe e , we assume ha missing in o ma ion is missing a andom and
ha he e is no sys ema ic bias.
Ano he limi a ion e e s o he aining indica o s used in his s udy. Regula aining
in es men s by companies and wo ke s’ pa icipa ion a es a e a a he ough measu e as
hey neglec di e ences in he leng h and impo ance o cou ses. As G een e al. (2000)
ha e shown o he B i ish case, he inc ease in aining pa icipa ion in ecen decades
has been accompanied by a dec ease in he olume o aining.
Finally, he da a do no allow us o con ol o he in luence o unobse ed indi idual
cha ac e is ics ha may a ec he ela ionship be ween ou independen and dependen
a iables. We know om exis ing esea ch ha his can lead o s a is ical dis o ions
(Becke and Schömann, 1996). The e ec s should he e o e be in e p e ed wi h his limi-
a ion in mind. We a emp ed o con ol o unobse ed ac o s a he company le el and
abo e, conside ing employee de elopmen , demand o skilled labou , wage de elop-
men o o he employee g oups and he ins i u ional company con ex .
O e all, he esul s o his s udy unde line he impo ance o di e ences in he shape
and egula ion o he in e nal labou ma ke . We made h ee cen al indings.
Fi s , we ound clea e idence suppo ing he main hesis o his a icle ha wo ke s’
chances o mo ing om semi-skilled o unskilled posi ions o highe (skilled) unc ions
a e shaped by he ins i u ional s uc u es o he company’s in e nal labou ma ke . Fo a
signi ican sha e o he wo ke s in ou sample, s aying wi h hei employe was associ-
a ed wi h conside ably lowe p omo ion p ospec s, hough wi h highe wages and job
secu i y.
Second, ega ding he shape o he in e nal labou ma ke s, we ound ha la ge com-
pany size, egula aining in es men s and egula ions a company le el we e associa ed
wi h signi ican ly be e chances o upwa d mobili y. The egula ion o he in e nal
labou ma ke ia o malised HR wo k p ocedu es – such as w i en a ge ag eemen s,
pe o mance app aisals, job desc ip ions, p ocedu es o illing acan posi ions o plans
o s a ing equi emen s – posi i ely in luenced he ca ee ad ancemen oppo uni ies
and wage de elopmen o wo ke s in semi-skilled and unskilled jobs. When he company
p omo ed in-company aining o his g oup o wo ke s on a egula basis, hei chances
o ad ancing o skilled occupa ional posi ions we e h ee imes highe . S uc u es o
employee ep esen a ion, in con as , we e nei he associa ed wi h be e p omo ion
chances no be e wage de elopmen o his g oup. Though wo ks councils a e o -
mally obliged o ep esen he in e es s o all wo ke s, hey migh ei he no be willing o
use hei ba gaining powe o be e p omo ion p ospec s o his g oup o wo ke s o no
able o p e ail agains managemen in e es s.
Thi d, ega ding he ole o he sec o al le el, ou s udy inds e idence o a con a y
ole o collec i e ag eemen s. On he one hand, hey os e ed a posi i e wage de elop-
men o wo ke s in unskilled o low-skilled jobs, p esumably by se ing s anda ds o
Wo schack and Sam leben 511
pay in ela ion o de ined wo k asks; on he o he hand, hey es ic ed p omo ion ajec-
o ies beyond he o malised oca ional aining sys em by gi ing g ea e weigh o o -
mal oca ional quali ica ions and aining measu es. In his espec , collec i e ag eemen s
un he isk o con ibu ing o he limi ed upwa d mobili y o wo ke s in semi-skilled and
unskilled jobs. Rega ding sec o al di e ences, ou s udy ound e idence ha ansi ions
om unskilled o semi-skilled o skilled jobs a e mo e di icul in he co e sec o s o he
Ge man sys em o indus ial ela ions (like manu ac u ing), p esumably due o he
s onge s anda disa ion and o malisa ion o quali ica ions and ca ee s.
Su p isingly, in companies wi h la ge aining in es men s ( aining p o ided o 50%
o mo e o semi-skilled and unskilled wo ke s), wo ke s’ chances o ca ee ad ancemen
we e educed and only inc eased i hey changed companies. These esul s a e consis en
wi h he p edic ions o he labou queue model (Thu ow, 1975) – long queues o p omo-
ion educe indi idual chances o upwa d mobili y and can be a oided by company
change.
Conclusions and implica ions o labou ma ke heo y and
policies
The esul s o his s udy unde line he impo ance o ins i u ional a angemen s a he
company and sec o al le el o he upwa d mobili y o wo ke s in semi-skilled o
unskilled jobs. This has been widely neglec ed in he li e a u e. How wo ke s in low-
skilled jobs ad ance in he labou ma ke is no only de e mined by di e en labou
ma ke segmen s (as s a ed by dual labou ma ke heo y) o indi idual cha ac e is ics (as
emphasised by neoclassical heo y), bu also shaped by ins i u ional egula ions and poli-
cies a company and sec o al le el (Appelbaum e al., 2010; K ings, 2020). S aying wi h
one employe was no ela ed o be e p omo ion chances o his g oup o wo ke s,
while he p esence o absence o inclusi e egula ions and collec i e in e es ep esen a-
ion made a signi ican di e ence.
A key inding o his s udy ega ds he ole o company policies and egula ions. In
line wi h a c i ical e iew and ex ension o dual labou ma ke heo y (see Appelbaum
e al., 2010; K ings, 2020; Lukac e al., 2019; Seo, 2021), ou s udy shows e idence ha
upwa d mobili y o wo ke s in unskilled and semi-skilled posi ions is shaped by la ge
company size and egula aining in es men s ocused on his g oup o wo ke s as well
as o malised HR policies co e ing he ields o pe o mance app aisals, job desc ip-
ions, p ocedu es o illing acan posi ions and plans o s a ing equi emen s. This
inding unde lines he need o conside and s udy ins i u ional cha ac e is ics o he in e -
nal labou ma ke mo e ca e ully. Mo eo e , possible ade-o s be ween p omo ion,
wages o employmen secu i y need o be conside ed (Ebene and Ehle , 2018; Seo,
2021) o e alua e hei impac on labou ma ke segmen a ion.
Rega ding sec o al di e ences and collec i e in e es ep esen a ion, ou s udy p o-
ides e idence o a e e se pa e n, in line wi h p edic ions om he compa a i e poli i-
cal economy li e a u e (Es é ez-Abe, 2005; Thelen, 2009: 482). In he highly egula ed
manu ac u ing sec o , ansi ions o skilled jobs a e mo e di icul o ealise, p esumably
due o he s onge s anda disa ion and o malisa ion o quali ica ions and ca ee
512 Wo k, Employmen and Socie y 39(2)
pa hways and hei en o cemen by collec i e ag eemen s and s ong unions. Rega ding
wage de elopmen , wo ke s in unskilled o semi-skilled jobs bene i when hey a e
employed in he manu ac u ing sec o o ha e ba gaining co e age. In his espec ,
s ong unions and collec i ely ag eed s anda ds ensu e he pa icipa ion o his g oup in
posi i e wage de elopmen bu es ic hei occupa ional upwa d mobili y.
A las inding conce ns he impo an ole o pa icipa ion in o mal oca ional ain-
ing. In line wi h p e ious esea ch (see, o example, Ehle , 2017), he esul s indica e
ha oppo uni ies o occupa ional s a us imp o emen s and wage inc eases o wo ke s
in low-skilled jobs depended subs an ially on pa icipa ion in mo e leng hy oca ional
aining, unde lining i s c ucial ole in b idging di e en labou ma ke segmen s in he
highly ins i u ionalised Ge man oca ional educa ion sys em.
Funding
The au ho s disclosed eceip o he ollowing inancial suppo o he esea ch, au ho ship, and/
o publica ion o his a icle: This esea ch was pa o he empi ical esea ch p ojec ‘The p o i-
sion and e ec s o company p o ided u he aining o he labou ma ke in eg a ion o low
skilled wo ke s’, unded by he Ge man Resea ch Founda ion (DFG). The open access publica ion
o his a icle was unded by he WZB Be lin Social Science Cen e ia he Sage Jou nals Online
Conso ium.
Supplemen al ma e ial
Supplemen al ma e ial o his a icle is a ailable online.
Re e ences
Allaa P, Bellmann L and Lebe U (2009) Company-p o ided u he aining in Ge many and
he Ne he lands. Empi ical Resea ch in Voca ional Educa ion and T aining 1(2): 103–121.
Appelbaum E, Bosch G and Gau ié J (2010) In oduc ion and o e iew. In: Gau ié J and Schmi J
(eds) Low-Wage Wo k in he Weal hy Wo ld. New Yo k, NY: Russell Sage Founda ion, 1–32.
A ulampalam W, Alison B and B yan M (2004) T aining in Eu ope. Jou nal o he Eu opean
Economic Associa ion 2(2–3): 346–360.
Asplund R (2005) The p o ision and e ec s o company aining: A b ie e iew o he li e a u e.
No dic Jou nal o Poli ical Economy 31: 47–73.
Bechmann S, Dahms V, Tsche sich N, e al. (2012) Fachk ä e und unbese z e S ellen in eine
al e nden Gesellscha . Nü nbe g: IAB-Fo schungsbe ich 13/2012.
Becke GS (1975) Human Capi al: A Theo e ical and Empi ical Analysis wi h Special Re e ence
o Educa ion. New Yo k, NY: Columbia Uni e si y P ess.
Becke R and Hecken AE (2009) Be u liche Wei e bildung – heo e ische Pe spek i en und
empi ische Be unde. In: Becke R (ed.) Leh buch de Bildungssoziologie. Wiesbaden: VS
Ve lag, 357–394.
Becke R and Schömann K (1996) Be u liche Wei e bildung und Einkommensdynamik. Kölne
Zei sch i ü Soziologie und Sozialpsychologie 48: 426–459.
Beich U, Walden G and He ge H (2004) Kos en und Nu zen de be ieblichen Be u sausbildung
in Deu schland. Biele eld: Bundesins i u ü Be u sbildung.
Bü mann M and Wiek J (2018) Ungele n e Fachk ä e. Fo male Un e quali ika ion in Deu schland.
Gü e sloh: Be elsmann S i ung.
Wo schack and Sam leben 513
BMAS and BMBF (2019) Na ionale Wei e bildungss a egie. Be lin: Bundesminis e ium ü
A bei und Soziales / das Bundesminis e ium ü Bildung und Fo schung.
BMBF (2015) Wei e bildungs e hal en in Deu schland 2014. AES 2014 T endbe ich . Bonn:
Bundesminis e ium ü Bildung und Fo schung.
BMBF (2021) Wei e bildungs e hal en in Deu schland 2020. E gebnisse des Adul Educa ion
Su ey – AES-T endbe ich . Be lin: Bundesminis e ium ü Bildung und Fo schung.
Bogai D, Wie höl e D, Buch T, e al. (2017) Digi alisie ung de A bei . Abschä zung de
Au oma isie ungspo enziale on Be u en in Be lin und B andenbu g. IAB-Regional 2/2017.
Nü nbe g: Ins i u ü A bei sma k - und Be u s o schung.
Boo h A and B yan M (2002) Who Pays o Gene al T aining? New E idence o B i ish Men and
Women. IZA Discussion Pape No. 486. Bonn: Ins i u e o Labo Economics (IZA).
Doellgas V, Lillie N and Pulignano V (eds) (2018) Recons uc ing Solida i y: Labou Unions,
P eca ious Wo k, and he Poli ics o Ins i u ional Change in Eu ope. Ox o d: Ox o d
Uni e si y P ess.
Dü sch M and S uck O (2014) A ypische Beschä igungen und be u liche Quali ika ions isiken
im E we bs e lau . Indus ielle Beziehungen 21(1): 58–77.
Ebene C and Ehle M (2018) Wei e bilden und Wei e kommen? Kölne Zei sch i ü Soziologie
und Sozialpsychologie 70: 213–235.
Ehle M (2017) Who bene i s om aining cou ses in Ge many? Mone a y e u ns o non- o mal
u he educa ion on a segmen ed labou ma ke . Eu opean Sociological Re iew 33: 436–448.
Eichho s W and Ma x P (2015) Non-S anda d Employmen in Pos -Indus ial Labou Ma ke s.
Chel enham: Edwa d Elga .
Ellgu h P, Kohau S and Mölle I (2014) The IAB es ablishmen panel – Me hodological essen ials
and da a quali y. Jou nal o Labou Ma ke Resea ch 47(1–2): 27–41.
Emmenegge P, Hause mann S, Palie B, e al. (eds) (2012) The Age o Dualiza ion: The Changing
Face o Inequali y in Deindus ializing Socie ies, online edn. In e na ional Policy Exchange
Se ies. Ox o d: Ox o d Academic.
Es é ez-Abe M (2005) Gende bias in skills and social policies: The a ie ies o capi alism pe -
spec i e on sex seg ega ion. Social Poli ics 12(2): 180–215.
E e sson M (2004) Fo mal on- he-job aining: A gende - yped expe ience and wage- ela ed
ad an age. Eu opean Sociological Re iew 20: 79–94.
G een F, Fels ead A, Mayhew K, e al. (2000) The impac o aining on labou mobili y: indi id-
ual and i m-le el e idence om B i ain. B i ish Jou nal o Indus ial Rela ions 38: 261–275.
G imshaw D, Fagan C, Hebson G, e al. (2017) A new labou ma ke segmen a ion app oach
o analysing inequali ies: In oduc ion and o e iew. In: G imshaw D, Fagan C, Hebson
G, e al. (eds) Making Wo k Mo e Equal: A New Labou Ma ke Segmen a ion App oach.
Manches e : Manches e Uni e si y P ess, 1–32.
Hansson B (2008) Job-Rela ed T aining and Bene i s o Indi iduals: A Re iew o E idence and
Explana ions. OECD Educa ion Wo king Pape s No. 19. Pa is: OECD.
Heining J, Jacobebbinghaus P, Scholz T, e al. (2012) Linked-Employe -Employee-Da en des IAB:
LIAB-Mo e -Modell 1993–2008 (LIAB MM 9308). FDZ Da en epo 01/2012. Nu embe g:
Ins i u e o Employmen Resea ch.
Köhle C, S uck O, G o hee M, e al. (2008) O ene und geschlossene Beschä igungssys eme.
Wiesbaden: Sp inge VS.
K ings T (2020) ‘Good’ bad jobs? The e olu ion o mig an low-wage employmen in Ge many
(1985–2015). Wo k, Employmen and Socie y 35(3): 527–544.
Lukac M, Doe linge N and Pulignano V (2019) De eloping a c oss-na ional compa a i e
amewo k o s udying labou ma ke segmen a ion. Social Indica o s Resea ch 145(1):
233–255.
514 Wo k, Employmen and Socie y 39(2)
Moh S, T ol sch K and Ge ha ds C (2016) Job asks and he pa icipa ion o low-skilled employ-
ees in employe -p o ided con inuing aining in Ge many. Jou nal o Educa ion and Wo k
29: 562–583.
OECD (2004) Imp o ing skills o mo e and be e jobs: Does aining make a di e ence? In:
OECD Employmen Ou look. Pa is: OECD.
Osiande C and S ephan G (2018) Un e welchen Bedingungen wü den sich Beschä ig e wei e -
bilden? E gebnisse eines ak o iellen Su eys. IAB-Discussion Pape No. 4/2018. Nü nbe g:
Ins i u ü A bei sma k - und Be u s o schung.
Pannenbe g M (1995) Wei e bildungsak i i ä en und E we bsbiog aphie. F ank u : Campus.
Pannenbe g M (2008) Indi iduelle E äge on Wei e bildung in KMU‘s und G oßbe ieben:
E idenz ü Wes deu schland. Soziale Fo sch i 57: 39–43.
Pa en D (1999) Wages and mobili y: The impac o employe -p o ided aining. Jou nal o Labo
Economics 17: 298–317.
P osse T (2016) Dualiza ion o libe aliza ion? In es iga ing p eca ious wo k in eigh Eu opean
coun ies. Wo k, Employmen and Socie y 30(6): 949–965.
Rahne S (2014) Zukun sau gabe Wei e bildung. Bonn: F ied ich-Ebe -S i ung.
Ramos CR and Ha is R (2012) T aining and i s bene i s o indi iduals. The In e na ional Jou nal
o Lea ning 18: 371–382.
Rigby M and Sanchis R (2021) The concep o skill and i s social cons uc ion. Eu opean Jou nal
o Voca ional T aining 37(1): 22–33.
Sande s J and de G ip A (2004) T aining, ask lexibili y and he employabili y o low-skilled
wo ke s. In e na ional Jou nal o Manpowe 25: 73–89.
Sengenbe ge W (1982) Beschä igungs- und a bei sma k poli ische An o de ungen an be iebli-
che Wei e bildung. In: Me ens D (ed.) Be u sbildungs o schung. Bei äge zu A bei sma k -
und Be u s o schung 66. Nü nbe g: Ins i u ü A bei sma k - und Be u s o schung, 253–278.
Seo H (2021) ‘Dual’ labou ma ke ? Pa e ns o segmen a ion in Eu opean labou ma ke s and
he a ie ies o p eca iousness. T ans e : Eu opean Re iew o Labou and Resea ch 27(4):
485–503.
Swank D, Ma in CJ and Thelen K (2008) Ins i u ional Change and he Poli ics o Social Solida i y
in Ad anced Indus ial Democ acies. In: Pape p esen ed a he 16 h In e na ional Con e ence
o Eu opeanis s, Council o Eu opean S udies, Chicago, IL, 6–8 Ma ch.
Thelen K (2009) Ins i u ional change in ad anced poli ical economies. B i ish Jou nal o Indus ial
Rela ions 47: 471–498.
Thelen K (2014) Va ie ies o Libe aliza ion and he New Poli ics o Social Solida i y. New Yo k,
NY: Camb idge Uni e si y P ess.
Thelen K and Busemeye M (2008) F om Collec i ism owa ds Segmen alism. MPI G Discussion
Pape 08/13. Cologne: Max Planck Ins i u e o he S udy o Socie ies.
Thu ow LC (1975) Gene a ing Inequali y. New Yo k, NY: Basic Books.
Wa hu s C, Tilly C and Ga a M (2017) A new social cons uc ion o skill. In: Buchanan J,
Finegold D, Mayhew K, e al. (eds) The Ox o d Handbook o Skills and T aining. Ox o d:
Ox o d Uni e si y P ess, 72–91.
Wo schack P (2020a) D i e s o aining pa icipa ion in low skilled jobs. The ole o ‘ oice’,
echnology, inno a ion and labo sho ages in Ge man companies. In e na ional Jou nal o
T aining and De elopmen 24(3): 245–264.
Wo schack P (2020b) When do companies ain low-skilled wo ke s? The ole o ins i u ional
a angemen s a he company and sec o al le el. B i ish Jou nal o Indus ial Rela ions 58(3):
587–616.
Zwick T (2003) P oduk i i ä se ek e inne be iebliche Wei e bildung. Pe sonal – Zei sch i ü
Human Ressou ce Managemen 55: 42–43.
Wo schack and Sam leben 515
Philip Wo schack is a esea che a he WZB Be lin Social Science Cen e and leade o he
esea ch g oup ‘Wo king wi h A i icial In elligence’ a he Weizenbaum Ins i u e o he
Ne wo ked Socie y. His esea ch in e es is ocused on issues o wo k, o ganisa ions, aining,
digi alisa ion and social inequali y. He holds a PhD in Social Sciences om he Uni e si y o
G oningen and led he esea ch p ojec ‘Company p o ided u he aining o he labou ma ke
in eg a ion o low-skilled wo ke s’, unded by he Ge man Resea ch Founda ion. Recen wo k has
been published in Eu opean Socie ies, Social Poli ics, he B i ish Jou nal o Indus ial Rela ions
and T ans e : Eu opean Re iew o Labou and Resea ch.
Clai e Sam leben is a p ojec manage a P ognos, an economic esea ch and consul ing company.
She holds a PhD in Sociology om he Humbold Uni e si y o Be lin and has wo ked in he
esea ch g oup ‘Time policy – how can social policy help o manage c i ical phases o e he li e
cou se?’ a he Ge man Ins i u e o Economic Resea ch. He esea ch in e es s include issues o
amily and labou sociology. Recen wo k on he di ision o ca e wo k and couples’ labou ma ke
pa icipa ion was published in Resea ch in Social S a i ica ion and Mobili y.
Da e submi ed July 2022
Da e accep ed June 2024
