Muelle , S e en; Neuschae e , Geo g
A icle — Published Ve sion
Wo ke Pa icipa ion in Decision‐making, Wo ke So ing,
and Fi m Pe o mance
Indus ial Rela ions: A Jou nal o Economy and Socie y
P o ided in Coope a ion wi h:
John Wiley & Sons
Sugges ed Ci a ion: Muelle , S e en; Neuschae e , Geo g (2021) : Wo ke Pa icipa ion in Decision‐
making, Wo ke So ing, and Fi m Pe o mance, Indus ial Rela ions: A Jou nal o Economy and
Socie y, ISSN 1468-232X, Wiley, Hoboken, NJ, Vol. 60, Iss. 4, pp. 436-478,
h ps://doi.o g/10.1111/i el.12288
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/284798
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 p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/
Wo ke Pa icipa ion in Decision-making, Wo ke
So ing, and Fi m Pe o mance
STEFFEN MUELLER and GEORG NEUSCHAEFFER
Wo ke pa icipa ion in decision-making is o en associa ed wi h high-wage and
high-p oduc i i y fi m s a egies. Using linked employe –employee da a o Ge -
many and wo ke fixed e ec s om a wo-way fixed-e ec s model o wages cap-
u ing obse ed and unobse ed wo ke quali y, we find ha plan s wi h o mal
wo ke pa icipa ion ia wo ks councils indeed employ highe quali y wo ke s.
We show ha wo ke quali y is al eady highe in plan s be o e council in oduc-
ion and u he inc eases a e he in oduc ion. Impo an ly, we co obo a e p e-
ious s udies by showing posi i e p oduc i i y and p ofi abili y e ec s e en a e
aking in o accoun wo ke so ing.
In oduc ion
Manda ed wo ke pa icipa ion in fi m decision-making is p esen in many
Eu opean coun ies o decades. Whe he employee pa icipa ion boos s p o-
duc i i y and d i es up wages has been discussed in ensi ely and is nowadays
inc easingly ele an agains he backg ound o he p oduc i i y slowdown and
alling labo sha es in na ional income. The Ge man model o plan -le el pa -
icipa ion ia wo ks councils has a ac ed pa icula in e es because o he
s ong legal igh s councils enjoy he e. S anda d economic heo y pe cei es
wo ks councils o be a labo ma ke ic ion gene a ing ad e se economic
e ec s (Jensen and Meckling 1979). Howe e , se e al o Ge man wo ks coun-
cils’legal igh s (discussed la e in mo e de ail) ha e he po en ial o inc ease
plan p oduc i i y di ec ly, o example, ia gene a ing collec i e oice, educ-
ing in o ma ion asymme ies be ween wo ke s and managemen , and os e ing
JEL codes: J5, J31, J24.
†
The au ho s’a filia ions a e, espec i ely, Halle Ins i u e o Economic Resea ch (IWH), Kleine M¨
a ke -
s aße 8, 06108 Halle, Ge many. O o- on-Gue icke-Uni e si y Magdebu g, Ge many and CESi o, Munich,
Ge many. E-mail: s e en.muelle[email p o ec ed]. Halle Ins i u e o Economic Resea ch (IWH), Halle, Ge -
many. E-mail: geo g.neuschae e @iwh-halle.de. We would like o hank Bo is Hi sch, Ma hias Me ens,
Jens Moh enweise , Claus Schnabel, and wo anonymous e e ees o use ul sugges ions. This esea ch has
been unded by he Eu opean Union’s Ho izon 2020 esea ch and inno a ion p og am, g an ag eemen No.
822390 (MICROPROD). Open Access unding enabled and o ganized by P ojek DEAL.
INDUSTRIAL RELATIONS, DOI: 10.1111/i el.12288. Vol. 60, No. 4 (Oc obe 2021). ©2021 The Au ho s.
Indus ial Rela ions published by Wiley Pe iodicals LLC on behal o Regen s o he Uni e si y o Cali o nia (RUC).
Published by Wiley Pe iodicals, Inc., 350 Main S ee , Malden, MA 02148, USA, and 9600 Ga sing on Road,
Ox o d, OX4 2DQ, UK.
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion-NonComme cial-NoDe i s
License, which pe mi s use and dis ibu ion in any medium, p o ided he o iginal wo k is p ope ly ci ed, he use
is non-comme cial and no modifica ions o adap a ions a e made.
436
us and longe - e m ela ions be ween hem. Exis ing empi ical esea ch
indeed demons a es ha council plan s ha e less employee u no e (Adam
2019; Addison, Schnabel and Wagne 2001; Hi sch, Schank and Schnabel
2010), pay highe wages (Addison e al. 2001; Hi sch and Muelle 2020), and
enjoy a p oduc i i y p emium (Muelle 2012; Muelle and S egmaie 2017).
Agains he backg ound o hese economically desi able e ec s, he con inued
decline in wo ks council co e age (Obe fich e and Schnabel 2019)
1
aises
conce ns abou p oduc i i y g ow h pe spec i es and wo ke s sha e in fi m
su plus.
Hi he o un ela ed o he wo ke pa icipa ion li e a u e, asso a i eness o
high-wage wo ke s o high-wage employe s has been documen ed in a numbe
o s udies.
2
As wo ks council plan s usually a e high-p oduc i i y, high-wage
employe s, a co e ques ion is whe he councils di ec ly inc ease hese ou -
comes o whe he council plan s employ wo ke s o highe quali y who will
inc ease p oduc i i y (see Bende e al. 2018) and ea n highe wages i espec-
i ely o wo ks council p esence. The wo ke code e mina ion li e a u e usually
a gues along he lines o he fi s scena io (e.g., Ji jahn and Smi h 2018; Muel-
le 2012) and places li le emphasis on po en ial sel -selec ion o high-quali y
wo ke s in o wo ks council plan s. Howe e , o mos wo ke s, going o a
high-paying employe o e ing s able employmen pe spec i es is a ac i e,
and hence, asso a i e ma ching o high-quali y wo ke s in o high-paying
wo ks council plan s is likely.
I wo ks councils a e a d i e o posi i e asso a i e ma ching, hen high-
wage, high-pe o mance plan s wi h wo ks councils would coexis wi h low-
wage, low-pe o mance plan s wi hou councils. This would no only imply
es ima ing spu ious p oduc i i y and wage gains om code e mina ion. I
would also sugges ha he legal manda e o councils con ibu es bo h o
be ween-plan wage inequali y (Ca d, Heining and Kline 2013; Hi sch and
Muelle 2020) and o p oduc i i y dispe sion ac oss plan s (Sy e son 2011).
1
An impo an ques ion is why wo ks council incidence declines despi e hese posi i e e ec s. F eeman
and Lazea (1995) a gue ha employe s figh agains p oduc i i y inc easing councils as long as he la e
de e io a e p ofi s. Wha is mo e, as soon as employe u ili y also depends on manage ial p e oga i es,
employe s may oppose e en p ofi -inc easing councils. Muelle and S egmaie (2020) eason ha employe
associa ions migh oppose p oduc i i y imp o ing wo ks councils as he la e ha e non-posi i e e ec s o
many small fi ms o ming he majo i y in employe associa ions.
2
This includes And ews e al. (2012) o Ge many, Bonhomme e al. (2019) using Swedish da a, and
Lopes de Melo (2018) o B azil. S udies applying wo-way fixed-e ec s models o wages as pionee ed in
Abowd e al. (1999) o en show e y small o e en nega i e asso a i e ma ching, o example Abowd e al.
(1999) o F ance and he Uni ed S a es. Howe e , he p ocedu e o Abowd e al. (1999) may unde es ima e
posi i e asso a i e ma ching due o limi ed mobili y bias (see And ews e al. 2008). Ca d e al. (2013) doc-
umen posi i e asso a i e ma ching o Ge many e en when using he me hod o Abowd e al. (1999).
Wo ks Councils, So ing and Pe o mance / 437
To analyze whe he so ing explains he p oduc i i y and wage e ec s o
wo ks councils, we a emp o imp o e on p io esea ch by u ilizing a sum-
ma y measu e o obse able and unobse able gene al human capi al compo-
nen s o wo ke s. Specifically, we use wo ke fixed e ec s om a wage
decomposi ion as pionee ed by Abowd, K ama z, and Ma golis (1999, hence-
o h AKM) and implemen ed by Ca d e al. (2013) o Ge many. In his
model, highe wo ke e ec s a e ewa ded highe ac oss all employe s, which
jus ifies labeling indi iduals wi h high AKM wo ke e ec s as high-quali y
wo ke s. Impo an ly, AKM wo ke e ec s cap u e all human capi al compo-
nen s ha a e in a ian in he ime span unde conside a ion and he e o e
include no only obse able human capi al a iables such as educa ion o ini-
ial age bu also unobse able componen s such as abili y.
3
Ou fi s con ibu-
ion will be o p esen e idence on he magni ude and he dynamics o so ing
by wo ks council exis ence.
P e ious s udies on p oduc i i y and wage e ec s o wo ks councils ypi-
cally y o con ol o wo ke quali y by means o (plan ) obse ables, o
example, by including he sha e o skilled wo ke s (Ji jahn and Muelle 2014;
Muelle 2012). To he ex en ha hese con ols do no ully cap u e unob-
se ed wo ke quali y di e ences, p e ious s udies may su e om an omi ed
a iable bias o unknown magni ude and ou abili y o con ol o unobse -
ables is a po en ially impo an con ibu ion o his li e a u e. We will also es
whe he he e is complemen a i y in labo p oduc i i y be ween wo ke pa ici-
pa ion and wo k o ce quali y, which is in o ma i e abou whe he such so ing
may imp o e alloca i e e ficiency.
Besides es ing whe he posi i e e ec s o wo ks councils on plan pe o -
mance and wages a e d i en by so ing, we also conside p ofi e ec s o see
whe he he ne e ec o code e mina ion on p oduc i i y and wages benefi s
employe s. In doing so, we examine o wha ex en he su plus gene a ed by
wo ks councils is sha ed wi h wo ke s and we he e o e p esen e idence on
how wo ke pa icipa ion in decision-making shapes he labo sha e a he
plan le el. To o e come any biases ha may s em om unobse ed plan
he e ogenei y, we apply an e en s udy amewo k and analyze wo ks council
in oduc ions in a wi hin-plan app oach and p o ide fi s e en s udy esul s
o wage and p ofi e ec s o councils.
4
We a gue ha he dynamics be o e
and a e council in oduc ion p o ide addi ional insigh s ega ding a causal
in e p e a ion o ou esul s.
3
A de ailed discussion o AKM wo ke e ec s will be p o ided in Sec ion “Da a and Empi ical S a -
egy”.
4
Muelle and S egmaie (2017) epo p oduc i i y e ec s in a simila se ing.
438 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
We will find ha council plan s indeed employ wo ke s o highe quali y
e en i a ich se o obse able plan cha ac e is ics is aken in o accoun .
Though some quali y di e ences exis al eady be o e he in oduc ion o a
wo ks council, hey widen as he council ma u es. We u he find ha he
sha e o high-quali y wo ke s s ongly inc eases plan s’labo p oduc i i y bu
ha he OLS es ima e o he wo ks council e ec declines only mode a ely by
one-fi h i AKM wo ke e ec s a e con olled o . In fixed-e ec s e en s udy
eg essions, he council e ec is unchanged when AKM wo ke e ec s a e
con olled o . This is good news o he alidi y o p e ious s udies as i
implies ha igno ing labo so ing, i a all, biased p e ious es ima es o labo
p oduc i i y e ec s o councils mode a ely upwa ds. We also find ha wo ks
council plan s pay highe wages, hough some inc ease in wages is al eady
p esen be o e he council’s in oduc ion. We show ha he su plus o igina ing
om he highe labo p oduc i i y o plan s wi h a wo ks council is sha ed by
employe s and wo ke s and find posi i e p ofi abili y e ec s in ou bo h OLS
and fixed-e ec s amewo ks. Wha is mo e, he p oduc i i y p emium o
high-quali y wo ke s is g ea e when a wo ks council is p esen , which sug-
ges s a complemen a i y be ween wo ke pa icipa ion and wo ke quali y. In
combina ion, ou fixed-e ec s e en s udy esul s show ha plan s in oducing
a wo ks council as compa ed o non-council plan s expe ience a u bulen ime
be o e in oduc ion wi h wo ke chu ning, s onge wage g ow h, and a p o-
duc i i y decline ha sha ply educe p ofi s p io o council in oduc ion. A e
council in oduc ion, wage g ow h fla ens and p oduc i i y g ow h se s in,
which allows council plan s o sus ain long- un p ofi abili y wi hin a high-
wage, high-p oduc i i y s a egy.
Ou pape is simila in spi i o Bende e al. (2018) who ocus on he ole
o managemen p ac ices and wo ke so ing on fi m p oduc i i y, a he han
o mal employee pa icipa ion. The main di e ence o Bende e al. (2018) is
ha we show how code e mina ion induces plan s o employ be e wo ke s,
ha is, acco ding o Bende e al. (2018), associa ed wi h he adop ion o supe-
io managemen p ac ices. In con as o Bende e al. (2018), we u ilize he
panel s uc u e o ou da a and show ha quali y upg ading indeed ollows
council in oduc ion. The main ake away will he e o e be ha an adequa ely
designed scheme o wo ke pa icipa ion in decision-making can shi plan s
in o an equilib ium wi h high wages and high p oduc i i y.
Ins i u ional Se ing, Theo y and Some Li e a u e
Regula o y amewo k and wo ke so ing. The Ge man sys em o indus-
ial ela ions es s on wo pilla s, ha is, plan -le el code e mina ion ia wo ks
Wo ks Councils, So ing and Pe o mance / 439

councils and sec o al collec i e wage ba gaining be ween unions and employe
associa ions.
5
The Wo ks Cons i u ion Ac (Be iebs e assungsgese z) equi es
wo ks councils o ac in he in e es o wo ke s and he plan and in a spi i o
mu ual us . The law u he codifies he ules o council elec ions and he
igh s elec ed councils ha e. Wo ke s o plan s wi h a leas fi e pe manen
employees ha e he igh o es ablish a council bu he e is no au oma ism o
do so. In ac , as o 2015 only 42 pe cen o wo ke s in Wes Ge many, which
will be he ocus o ou analysis, wo ked in he 9 pe cen o eligible plan s
ha ha e a wo ks council (Ellgu h and Kohau 2016).
6
The Wo ks Cons i u ion Ac g an s councils se e al in o ma ion and consul-
a ion igh s and addi ionally defines opics whe e councils a e able o block
decisions ( e o igh s) o ha e he igh o code e mine social ma e s. In o ma-
ion igh s, o ins ance, include he igh o ge access o in o ma ion on he
plan ’s economic and financial si ua ion. These igh s pu councils in he posi-
ion o e i y managemen p o ided in o ma ion and, hus, po en ially lead o
a mo e c edible op-down communica ion. By educing in o ma ion asymme-
ies be ween wo ke s and he employe , in o ma ion igh s may, o ins ance,
p e en ine ficien plan closu e (F eeman and Lazea 1995). Wo ks councils
ha e o be in o med and consul ed i he employe plans majo changes in he
wo k en i onmen o he p oduc ion p ocess. On he one hand, consul a ion-
induced decision delay migh be cos ly, bu on he o he hand, i managed
app op ia ely, he consul a ion p ocess add esses po en ial ea s o wo ke s and
esul s in a well-in o med wo k o ce being mo e commi ed o desi ed
changes.
7
Wo ks councils’code e mina ion igh s a e s onges in social ma e s. Fo
ins ance, i a council o mally disag ees wi h an indi idual dismissal his dis-
missal u ns oid un il a labo cou finally decides he ma e . Fi ing cos s
hus inc ease o employe s, and his may well ha e implica ions on
5
Fo excellen heo e ical discussions on non-union wo ke ep esen a ion and he Ge man expe ience,
we e e o Addison (2009) and Ji jahn and Smi h (2018).
6
Why only a small and declining sha e o eligible plan s has a council (Obe fich e and Schnabel 2019)
is no ully unde s ood. Employe s a e p ohibi ed o in e e e wi h wo ks council elec ions and e en ha e o
bea he cos s o unning he elec ion. Once elec ed, councilo s enjoy e y s ong employmen p o ec ion.
Because o his, and because ime spen on wo k as a wo ks councilo coun s as egula wo king ime, he
nonexis ence o councils in many eligible plan s poin s o addi ional cos s po en ial councilo s ace. This
cos may, o ins ance, include he cos s o posi ioning onesel as a wo ks councilo , while many employe s
ha e ese a ions agains code e mina ion (Muelle and S egmaie 2020) and he cos s o ac i ely o ganizing
a join posi ion o wo ke s, ep esen ing hei in e es s, and being esponsible o he nego ia ion ou comes.
7
The link be ween council exis ence and inno a i e ac i i y has been analyzed in Schnabel and Wagne
(1994), Addison and Wagne (1997), and Addison e al. (2001). Nei he o hese s udies ound any s a is i-
cally significan ela ionship. In e es ingly, Ji jahn and K a (2011) find a posi i e link wi h inc emen al p o-
duc inno a ions bu no wi h d as ic inno a ions.
440 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
p oduc i i y and so ing. Inc eased fi ing cos s may, on he one hand, de e io-
a e p oduc i i y by educing incen i es o wo k ha d (Addison e al. 2001, p.
671) bu , on he o he hand, le bo h sides ake a longe - e m iew on he
employmen ela ionship, which incen i izes indi idual wo ke s o ca e abou
he economic iabili y o hei plan . Employe s may eac o inc eased fi ing
cos s by in es ing in sc eening ac i i ies when hi ing new wo ke s o in oduce
high-pe o mance wo k p ac ices such as pe o mance pay (Lazea 2000),
which in u n should imp o e hei abili y o iden i y and a ac high-
p oduc i i y wo ke s. When laying o wo ke s ge s expensi e, employe s in
code e mined plan s may p o ide addi ional aining measu es (S egmaie
2012) o upg ade he skills o hei incumben wo ke s o allow he la e o
compe e wi h well- ained labo ma ke en an s (Janssen and Moh enweise
2018).
Wha is mo e, he s anda d “collec i e oice”a gumen can be made also
o wo kplace ep esen a ion ia wo ks councils. “Collec i e oice”(F eeman
1976) as opposed o “exi oice”(Hi schman 1970) emphasizes ha wo ke
ep esen a ion a he wo kplace gi es dissa isfied wo ke s a chance o anony-
mously exp ess hei dissa is ac ion wi hou ha ing o ea sanc ions by he
employe . This may p e en hese wo ke s om qui ing hei jobs (o om
educing e o wi hou qui ing o mally), and i p o ides employe s wi h
mo e in o ma ion abou wo ke p e e ences han exi oice would do.
Bo h he fi ing cos a gumen and he collec i e oice a gumen imply
educed wo ke u no e in code e mined plan s. Using plan -le el da a, F ick
(1996) finds ha wo ks council exis ence is ela ed o ewe qui s and, among
o he s, Addison e al. (2001), F ick and M¨
olle (2003), P ei e (2011) and
G und, Ma in and Schmi (2016) confi m ha u no e is educed. Whe he
hese a e indeed di ec “collec i e oice”e ec s o whe he hey a e a he
en -seeking e ec s is analyzed by Hi sch e al. (2010) and Adam (2019).
U ilizing employe –employee da a, Hi sch e al. (2010) find oice e ec s only
o a subg oup o low enu e wo ke s. Adam (2019) eso s on plan -le el da a
and exploi s a change in he legal amewo k wi hin a di e ence-in-di e ences
se ing and finds s ong oice e ec s as he sou ce o educed u no e . To
sum up, he li e a u e almos uni o mly finds educed employee u no e and
some ole o “collec i e oice”in explaining i .
On op o enjoying a s able job and s onge legal igh s in he wo kplace,
one o he main a gumen s o wo ke s o mo e o wo ks council plan s is ha
he la e pay wage p emia o hei wo ke s. This is documen ed in Hi sch and
Muelle (2020) who show ha councils a e associa ed wi h highe employe
wage p emia e en condi ional on plan s’quasi- en s and accoun ing o wo ke
so ing.
Wo ks Councils, So ing and Pe o mance / 441
Ha ing discussed why high-quali y wo ke s ma ch wi h councils fi ms,
ano he channel os e ing asso a i e ma ching migh come om highe quali y
wo ke s’incen i es o es ablish a council o p o ec hei quasi- en s. Ji jahn
(2009) a gues ha wo ke s who in es ed in o hei human capi al will ha e a
s ong incen i e o p o ec hei quasi- en s by ounding a council in fi ms wi h
de e io a ing economic pe o mance. Besides documen ing a highe likelihood
o council adop ion in poo ly pe o ming fi ms, Ji jahn (2009) also shows ha
plan s wi h a highe ac ion o skilled blue-colla wo ke s a e mo e likely o
ound a council. Whe eas Ji jahn and Moh enweise (2016) and Obe fich ne
(2019) also epo a highe likelihood o council in oduc ions o plan s wi h
high-skilled wo ke s, Addison e al. (2013) and Moh enweise , Ma ginson and
Backes-Gellne (2012) do no find suppo o his no ion. We a e no awa e o
any s udy on council in oduc ions ha inco po a es measu es o wo ke qual-
i y ha go beyond obse able skill le els.
Collec i e wage ba gaining be ween unions and employe associa ions o ms
he second pilla o indus ial ela ions in Ge many. In 2015, 59 (31) pe cen
o wo ke s (plan s) we e co e ed by collec i e ag eemen s in Wes Ge many
(Ellgu h and Kohau 2016). The Wo ks Cons i u ion Ac cla ifies he ela ion-
ship be ween wo ks councils and unions by s ipula ing ha councils a e no
allowed o in e e e wi h union wage se ing and a e no allowed o call
s ikes. Al hough o mally independen o each o he , wo ks councils and
unions ha e close ies, o example, p o iding wo ks councilo s wi h esou ces
and councils ec ui ing new union membe s a he shop floo (Beh ens 2009).
F eeman and Lazea (1995) a gue ha he exis ence o sec o -le el wage ba -
gaining should inc ease he p oduc i i y e ec o councils because councils a e
hen less engaged in dis ibu ional conflic s and ca e mo e abou inc easing he
o e all pie o be sha ed be ween wo ke s and he employe .
Wo ks councils and plan and wo ke ou comes. As he li e a u e on he
economic consequences o wo ks councils has no sys ema ically examined
(unobse ed) wo ke quali y so ing by council s a us, he subsequen li e a u e
e iew ocuses on he li e a u e on economic consequences o Ge man wo ks
councils in gene al.
P oduc i i y. The empi ical economic li e a u e on he p oduc i i y e ec
o wo ks councils s a ed in he 1980s. While ea ly s udies had o ely on e y
small samples and es ima ed nega i e council e ec s (Fi zRoy and K a
1987), la e s udies we e able o u ilize la ge-scale plan -le el da a. As a wo k-
ho se model, hese s udies employed p oduc ion unc ion es ima ions in which
a council dummy indica es he ce e is pa ibus p oduc i i y ad an age/disad an-
age o wo ks council exis ence. Council coe ficien s om OLS es ima ions
442 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
ange om 15 pe cen in Wol and Zwick (2002) and 18 pe cen in Muelle
(2015) o 25 pe cen in Addison, Schank, Schnabel and Wagne (2006) and
e en 30 pe cen in F ick and M¨
olle (2003). Though hese s udies usually con-
ol o he ac ion o skilled c a sman in he wo k o ce (and some imes also
o he sha e o uni e si y g adua es), hey we e no able o con ol o addi-
ional human capi al componen s such as wo ke expe ience o unobse ed
abili y. Muelle (2012, 2015) analyzes he council’s p oduc i i y e ec and
con ol o he ac ions o skilled wo ke s, app en ices, and pa - ime wo ke s
in he wo k o ce and o he capi al s ock. Muelle (2012) combines a GMM-
SYS p oduc ion unc ion es ima ion wi h an endogenous swi ching eg ession
and finds a p oduc i i y e ec o abou 7 pe cen in he manu ac u ing sec o ,
and Muelle (2015) employs ecen e ed influence unc ion echniques (Fi po,
Fo in and Lemieux 2009) and epo s ha he council e ec is highe in less
p oduc i e plan s. Fu he mo e, F eeman and Lazea ’s (1995) hypo hesis o a
mode a ing e ec o sec o -le el wage ba gaining on he p oduc i i y e ec o
councils has ecei ed s ong suppo in empi ical wo k (e.g., B ¨
andle 2017;
H¨
uble and Ji jahn 2003; Ji jahn and Muelle 2014).
One majo issue ha has long been un esol ed is wo ks council endogenei y
due o unobse ed plan he e ogenei y as a sou ce o bias in wo ks council
p oduc i i y es ima es. The main di ficul y wi h unobse ed he e ogenei y is
ha wo ks council s a us does a ely change wi hin plan s o e ime, which
makes i ha d o de ec s a is ically significan e idence in any fixed-e ec s o
fi s -di e ence es ima ion s a egy. Ea ly a emp s o use fixed-e ec s es ima-
o s indeed yielded insignifican p oduc i i y e ec s (Addison, Schnabel and
Wagne 2004).
8
Howe e , wi h much mo e obse a ions a hand, Muelle and
S egmaie (2017) ecen ly showed wi hin a fixed-e ec s e en s udy app oach
ha wo ks councils a e associa ed wi h declining p oduc i i y p io o council
in oduc ion and ha p oduc i i y g ow h ou paced ha o non-council plan s
a e an in oduc ion pe iod o abou fi e yea s, leading o a subs an ial p o-
duc i i y p emium o council plan s in he long un.
9
The p e-in oduc ion
decline in p oduc i i y is in line wi h he findings in K a and Lang (2008),
Ji jahn (2009), and Moh enweise e al. (2012) who find ha councils a e
in oduced in plan s acing ad e se condi ions, a finding ha has epea edly
been used o a gue ha con en ional es ima es o p oduc i i y e ec s o wo ks
8
H¨
uble and Ji jahn (2003) and Muelle (2012) aim on ackling council endogenei y by using endoge-
nous swi ching eg ession models. Bo h find posi i e e ec s bu , as hese models ei he iden i y e ec s
exclusi ely ia assump ions on he join dis ibu ion o e o e ms (H¨
uble and Ji jahn 2003) o , addi ion-
ally, by an exclusion es ic ion ha may o may no hold (Muelle 2012), he ma e o sel -selec ion can be
conside ed as being s ill un esol ed.
9
Ji jahn e al. (2011) epo a hump-shaped link be ween council age and p oduc i i y in hei OLS se -
ing.
Wo ks Councils, So ing and Pe o mance / 443
plan s. In e es ingly, he sha e o skilled wo ke s, which is a defini ion cap u -
ing a ela i ely b oad skill se (see Appendix A), is e y simila ac oss bo h
g oups o plan s indica ing ha AKM wo ke e ec s indeed con ey di e en
in o ma ion and dis inguishes be e be ween wo ke s o di e en quali y.
Toge he wi h he esul s on p oduc i i y, p ofi s, and wages, he desc ip i e
analysis he e o e poin s o s ong asso a i e ma ching o high-wage wo ke s
o high-wage, high-p oduc i i y plan s.
Columns (3) and (4) summa ize he ou comes o plan s be o e he in oduc-
ion o a wo ks council. Compa ed wi h non-council plan s, he 67 plan s
TABLE 1
SUMMARY STATISTICS
Va iable
Wo ks council No wo ks council
Yea s be o e council in oduc ion
less han 3 a leas 3
Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Log(labo p oduc i i y) 11.280 (0.606) 10.905 (0.663) 11.105 (0.685) 11.351 (0.721)
Log(wage bill pe wo ke ) 10.386 (0.397) 10.006 (0.527) 10.244 (0.473) 10.218 (0.528)
Log(p ofi pe wo ke ) 10.281 (1.232) 9.894 (1.275) 10.009 (1.346) 10.608 (1.164)
Log(employmen ) 5.368 (1.201) 3.141 (1.040) 4.225 (1.088) 3.981 (0.984)
Log(capi al in ensi y) 11.042 (1.234) 10.509 (1.246) 10.488 (1.875) 10.620 (1.202)
Collec i e ba gaining 0.783 (0.412) 0.341 (0.474) 0.414 (0.494) 0.340 (0.476)
Wo ke quali y 0.292 (0.700) −0.144 (1.090) 0.284 (0.884) 0.262 (0.923)
Skilled employees as
sha e o all employees 0.674 (0.250) 0.643 (0.252) 0.689 (0.279) 0.670 (0.273)
Pa - ime employees as
sha e o all employees 0.110 (0.162) 0.209 (0.205) 0.146 (0.206) 0.146 (0.218)
App en ices as
sha e o all employees 0.041 (0.039) 0.051 (0.073) 0.043 (0.050) 0.044 (0.051)
Female employees
sha e o all employees 0.275 (0.215) 0.374 (0.262) 0.317 (0.259) 0.332 (0.243)
Chu ning a e 0.041 (0.066) 0.059 (0.267) 0.068 (0.134) 0.064 (0.118)
Expo e 0.693 (0.461) 0.363 (0.481) 0.468 (0.500) 0.651 (0.479)
Single plan 0.509 (0.500) 0.835 (0.371) 0.516 (0.501) 0.651 (0.479)
Technical s a e o machine y
excellen 0.176 (0.381) 0.220 (0.414) 0.253 (0.436) 0.236 (0.427)
good 0.514 (0.499) 0.503 (0.500) 0.414 (0.494) 0.575 (0.497)
ai 0.275 (0.447) 0.256 (0.436) 0.306 (0.462) –†
poo 0.034 (0.181) 0.022 (0.145) –† –†
A e age wo ke age 41.961 (3.489) 41.132 (5.501) 39.865 (3.992) 39.308 (3.978)
Uni e si y deg ee
sha e o all employees 0.082 (0.122) 0.063 (0.135) 0.095 (0.177) 0.103 (0.210)
N7467 15,109 186 106
No es: LIAB c oss-sec ional model, 1998–2016, Wes Ge many. Summa y o 22,576 plan -yea obse a ions. Wo ke qual-
i y is he mean o he AKM wo ke e ec s (α
i
) a he plan le el (as desc ibed in Sec ion “Da a and Empi ical S a egy”)
s anda dized wi h a mean o ze o and a s anda d de ia ion o one.
–†The alues a e no shown due o easons o da a p o ec ion.
450 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER

in oducing a council ha e highe labo p oduc i i y, wages, and wo ke qual-
i y e en be o e he in oduc ion.
20
Thei ou comes a e, howe e , wo se han
hose o council plan s, which indica es ha wo ks council in oduc ion may
u he imp o e ou comes. Hence, ou esul s show desc ip i ely ha plan s
wi h high pe o mance, somewha highe wo ke chu ning, and high wo ke
quali y seem o be mo e likely o in oduce a council and ha pe o mance
and wages inc ease a e council in oduc ion whe eas chu ning dec eases. To
sc u inize hese esul s, we la e show he dynamics be o e and a e he coun-
cil in oduc ion in a mul i a ia e fixed-e ec s e en s udy se ing.
Wo ke so ing. When analyzing wo ke so ing, he s anda dized plan -
le el a e age o he AKM wo ke e ec (αj ) becomes he dependen a iable
in model (2). The co esponding OLS eg ession esul s a e displayed in
Table 2. Wo ks council exis ence en e s posi i ely and significan ly in all spec-
ifica ions. Omi ing he sha e o skilled wo ke s (column 1) yields a wo ks
council coe ficien o 0.194, implying ha wo ke quali y is highe by nea ly
one-fi h o a s anda d de ia ion. Con olling o he ac ion o skilled wo ke s
educes he coe ficien o 0.157 (column 2). Remembe ha he AKM wo ke
e ec cap u es also obse able human capi al componen s embodied in age
and o mal educa ion (see Sec ion Da a). Including in he eg ession bo h a e -
age wo ke age and he sha e o wo ke s ha ing an uni e si y deg ee educes
he council coe ficien o 0.115 (column 4). Hence, including bo h obse able
human capi al componen s does only accoun o a modes ac ion o he
wo ke quali y e ec .
The e en s udy esul s o he a e age AKM wo ke e ec a he plan le el
a e depic ed in Figu e 1. I shows he coe ficien s o he ela i e ime dum-
mies
21
wi h hei 90% confidence in e als, whe e he h ee yea s be o e in o-
duc ion (−3 o−1) se e as base ca ego y. In ou baseline specifica ion
(Figu e 1A), wo ke quali y ises by 0.159 s anda d de ia ions om he p e-
in oduc ion pe iod o a wo ks council age o mo e han eigh yea s. The
insignifican p e-e en end suppo s he conclusion ha council in oduc ion
inc eases wo ke quali y as opposed o a na a i e whe e councils a e in o-
duced in plan s ha would ha e upg aded wo ke quali y anyway. Ou esul s
emain unchanged when we omi he con ol a iables (Figu e 1A) so ha we
conclude ha ou esul s a e no a ec ed by any issue ha migh a ise om
20
Mo e han h ee yea s be o e council in oduc ion, he e a e 106 plan -yea obse a ions, in he
3, 1
½
– ela i e ime in e al 186; in he 0, 2
½
–in e al 183; in he 3, 5
½
–in e al 136; and in he 6, 8
½
–
in e al 126, and nine yea s a e council in oduc ion, we ha e 155 plan -yea obse a ions.
21
The six ela i e ime in e als a e imek
cτ(k∈∞,4½;3, 1½;0, 2½;3, 5½;6, 8½;9, ∞½
g
)
Wo ks Councils, So ing and Pe o mance / 451
con olling o pos - ea men ealiza ions o he con ol a iables (some imes
called “bad con ol”p oblem).
Upg ading along ime-in a ian obse able wo ke cha ac e is ics is one pos-
sible explana ion o wo ke quali y imp o emen s a e council in oduc ion.
TABLE 2
WORKER QUALITY, OLS REGRESSIONS
(1) (2) (3) (4)
Wo ks council 0.194*** 0.157*** 0.142*** 0.115***
(0.028) (0.026) (0.032) (0.025)
Skilled employees 0.857*** 0.858*** 0.700***
(0.043) (0.043) (0.042)
Collec i e ba gaining 0.039*0.011 0.004 0.007
(0.023) (0.022) (0.027) (0.021)
Wo ks council ×collec i e ba gaining 0.026
(0.039)
Log(capi al in ensi y) 0.038*** 0.028*** 0.028*** 0.028***
(0.010) (0.009) (0.009) (0.009)
Expo e 0.130*** 0.135*** 0.134*** 0.090***
(0.027) (0.025) (0.025) (0.024)
Single plan −0.166*** −0.133*** −0.133*** −0.109***
(0.023) (0.022) (0.022) (0.021)
Technical s a e =good −0.043*−0.026 −0.026 −0.036*
(0.022) (0.022) (0.022) (0.021)
Technical s a e = ai −0.124*** −0.079*** −0.079*** −0.094***
(0.027) (0.026) (0.026) (0.025)
Technical s a e =poo −0.155*** −0.095** −0.096** −0.122***
(0.047) (0.045) (0.045) (0.045)
Pa - ime employees −0.224*** −0.016 −0.016 −0.040
(0.086) (0.083) (0.083) (0.082)
App en ices −0.448** −0.200 −0.195 0.448**
(0.190) (0.182) (0.181) (0.192)
Female employees −0.682*** −0.559*** −0.558*** −0.558***
(0.071) (0.066) (0.066) (0.063)
Chu ning a e −0.087 −0.056*−0.056*−0.041*
(0.053) (0.032) (0.032) (0.024)
A e age wo ke age 0.015***
(0.003)
Uni e si y deg ee 1.834***
(0.105)
Cons an 0.038 −0.655 −0.651 −1.034**
(0.635) (0.615) (0.615) (0.524)
R
2
0.304 0.339 0.339 0.377
N22,576 22,576 22,576 22,576
No es: LIAB c oss-sec ional model, 1998–2016, Wes Ge many, OLS sample. ***/**/*deno es s a is ical significance a
he 1%/5%/10% le el. The dependen a iable is he mean o he AKM wo ke e ec s (α
i
) a he plan le el (as
desc ibed in Sec ion “Da a and Empi ical S a egy”) s anda dized wi h a mean o ze o and a s anda d de ia ion o one.
Repo ed numbe s a e coe ficien s om OLS eg essions wi h s anda d e o s clus e ed a he plan le el in pa en heses.
Fu he co a ia es included in all specifica ions a e 7 plan size dummies, 8 ede al s a e dummies, 37 wo-digi sec o
dummies, and 18 ime dummies.
452 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 6 o 8 ≥ 9
ela i e ime in e als
baseline wi hou con ol a iables
90% con idence in e als
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 6 o 8 ≥ 9
ela i e ime in e als
baseline condi ional on wo ke age
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 6 o 8 ≥ 9
ela i e ime in e als
baseline condi ional on uni e si y deg ee
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 6 o 8 ≥ 9
ela i e ime in e als
baseline condi ional on age and educa ion
FIGURE 1
WORKER QUALITY,EVENT STUDY.
No es: LIAB c oss-sec ional model, 1998–2016, Wes Ge many, e en s udy sample (141,098
plan -yea -coho obse a ions). Wo ks council in oduc ions be ween 1998 and 2016. This
figu e shows he mean ou come o wo ke quali y ela i e o he p e-in oduc ion pe iod o he
wo ks council and ne o he e olu ion in he con ol g oup. Wo ke quali y is measu ed as he
mean o he AKM wo ke e ec s (α
i
) a he plan le el (as desc ibed in Sec ion “Da a and
Empi ical S a egy”) s anda dized wi h a mean o ze o and a s anda d de ia ion o one. The
e en (yea =0 o 2) is he in oduc ion pe iod o he wo ks council, and ela i e ime (in
yea s) is depic ed a he ho izon al axis. As specified in equa ion (3), he eg ession includes
con ols o collec i e wage ag eemen p esence, capi al in ensi y, expo s a us, single-plan
s a us, he s a e o echnical machine y, he sha e o skilled employees, pa - ime wo ke s,
app en ices and women o all employees, 7 plan size dummies, and plan -coho fixed e ec s.
Fo panel A: No con ols o he han specified in equa ion (3) a e used. Fo panel B: holding
cons an wo ke age a he plan le el. Fo Panel C: holding cons an he sha e o uni e si y
g adua es. Fo Panel D: holding cons an age and he sha e o uni e si y g adua es. The 90%
confidence in e als a e shown using s anda d e o s clus e ed a he plan le el [Colo figu e
can be iewed a wileyonlinelib a y.com]
Wo ks Councils, So ing and Pe o mance / 453
We add a e age wo ke age o he e en s udy, bu he pos -e en coe ficien s
s ay unchanged (Figu e 1B). The council coe ficien s do no change a lo
ei he , when we con ol o he sha e o uni e si y g adua es alone (Figu e 1C)
o join ly wi h a e age wo ke age (Figu e 1D). We conclude ha he inc ease
in wo ke quali y is d i en by an inc ease in unobse ed componen s o he
AKM wo ke e ec .
22
I is in e es ing o unde s and whe he he highe a e age wo ke quali y
esul s om wo ke chu ning o skill upg ades o incumben wo ke s. Using
he OLS sample, we es ima e model 2 explaining he a e age AKM wo ke
e ec s o ei he joining, lea ing, and s aying wo ke s, espec i ely. We also
analyzed di ec ly he di e ence in he AKM e ec s o joine s e sus (lagged)
lea e s and he fi s di e ence in he AKM e ec s o s aye s.
23
As joine s (lea-
e s) canno be iden ified o a plan ’sfi s (las ) obse a ion, sample size
dec eases subs an ially. We find ha he AKM wo ke e ec s o joining, s ay-
ing, and lea ing wo ke s a e highe in wo ks council plan s. Impo an ly, we
also find ha he di e ence in he AKM e ec s be ween joine s and lea e s is
mo e posi i e and s a is ically significan in wo ks council plan s (Table 3, col-
umn 4), whe eas we find no s a is ically significan di e ence be ween council
and non-council plan s ega ding he change in he AKM e ec s o s aye s
(Table 3, column 5). These esul s a o he no ion ha wo ke quali y
imp o emen s in council plan s a e a he d i en by wo ke chu ning han by
quali y imp o emen s o incumben wo ke s.
24
Summing up, we find ha wo ke quali y is highe in plan s wi h a wo ks
council han in plan s wi hou , ha his di e ence is pa ly al eady p esen
be o e council in oduc ion (see Table 2), ha i inc eases u he a e he
council is in oduced, and ha his inc ease is bes explained by wo ke chu n-
ing. Ou e en s udies u he show ha imp o emen s in unobse ed wo ke
quali y a he han changes in wo ke s’ o mal educa ion o age d i e quali y
imp o emen s.
22
In Appendix B, we p esen u he obus ness checks o selec i e panel a i ion including only obse -
a ions o which we obse e he council in oduc ion om he panel s uc u e o ou da a. We find ha ou
main esul s a e unchanged.
23
We weigh he a e ages o joining, lea ing, and s aying wo ke s wi h hei sha e in plan employmen
o accoun o hei ela i e impo ance o he plan ’s a e age AKM e ec . This is c ucial as chu ning may
no only a ec he a e age quali y wi hin hese g oups o wo ke s bu also he weigh wi h which ei he o
he h ee g oups en e he plan a e age.
24
P esumably due o he sha ply educed sample size, we ound no clea e idence o ei he he chu ning
channel no o skill upg ades o incumben wo ke s when we apply fixed-e ec s e en s udy eg essions.
454 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
TABLE 3
WORKER QUALITY OF JOINING,LEAVING,AND STAYING WORKERS, OLS REGRESSIONS
Wo ke quali y Di e ence o wo ke quali y
Joining
wo ke s
Lea ing
wo ke s
S aying
wo ke s
Joining s.
lea ing
S aying s. s aying o
1
(1) (2) (3) (4) (5)
Wo ks council 0.129*** 0.073*** 0.136*** 0.058** −0.032
(0.025) (0.026) (0.038) (0.024) (0.027)
Skilled employees 0.498*** 0.469*** 0.888*** 0.037 0.050
(0.053) (0.052) (0.054) (0.048) (0.048)
Collec i e
ba gaining
0.056** 0.032 0.002 0.025 −0.032
(0.024) (0.022) (0.027) (0.023) (0.024)
Log(capi al
in ensi y)
0.010 0.021*0.034*** −0.012 0.001
(0.011) (0.011) (0.011) (0.010) (0.010)
Expo e 0.028 0.032 0.083*** −0.004 −0.018
(0.027) (0.024) (0.030) (0.025) (0.024)
Single plan −0.071*** −0.041*−0.127*** −0.032 0.008
(0.021) (0.021) (0.026) (0.020) (0.021)
Technical s a e =
good
−0.025 −0.046*−0.048*0.021 −0.026
(0.025) (0.025) (0.026) (0.027) (0.028)
Technical s a e =
ai
−0.054*−0.031 −0.085*** −0.024 −0.020
(0.029) (0.029) (0.031) (0.030) (0.032)
Technical s a e =
poo
−0.057 −0.030 −0.084 −0.029 −0.003
(0.069) (0.062) (0.062) (0.076) (0.054)
Pa - ime
employees
−0.505*** −0.304*** 0.179 −0.212*0.066
(0.130) (0.117) (0.111) (0.119) (0.096)
App en ices 0.632** 0.137 0.940*** 0.510 −0.284
(0.281) (0.298) (0.269) (0.322) (0.261)
Female employees −0.363*** −0.484*** −0.703*** 0.116 −0.042
(0.083) (0.083) (0.082) (0.081) (0.071)
Chu ning a e −0.680*−0.708*** 0.068 0.017 0.126
(0.352) (0.229) (0.064) (0.170) (0.226)
A e age wo ke
age
0.024*** 0.021*** 0.008** 0.003 −0.013***
(0.005) (0.004) (0.004) (0.004) (0.004)
Uni e si y deg ee 1.301*** 1.151*** 2.712*** 0.173 0.288**
(0.149) (0.145) (0.169) (0.129) (0.128)
Cons an −2.205*** −1.870*** −2.041*** −0.374 0.735***
(0.330) (0.288) (0.284) (0.289) (0.261)
R
2
0.308 0.306 0.511 0.015 0.083
N10,317 10,317 10,317 10,317 10,317
No es: LIAB c oss-sec ional model, 1998–2016, Wes Ge many, OLS sample. ***/**/*deno es s a is ical significance a
he 1%/5%/10% le el. The dependen a iables o column 1 o 3 a e he means o he AKM wo ke e ec s (α
i
) a he
plan le el (as desc ibed in Sec ion “Da a and Empi ical S a egy”) o joining, lea ing, and s aying wo ke s, espec i ely,
weigh ed wi h hei sha es in plan employmen . The dependen a iables o column 4 and 5 a e he di e ence o he
weigh ed means o joining and lea ing wo ke s and he fi s di e ence o he weigh ed means o s aying wo ke s, espec-
i ely. The dependen a iables a e s anda dized wi h a mean o ze o and a s anda d de ia ion o one. Repo ed numbe s
a e coe ficien s om OLS eg essions wi h s anda d e o s clus e ed a he plan le el in pa en heses. Fu he co a ia es
included in all specifica ions a e 7 plan size dummies, 8 ede al s a e dummies, 37 wo-digi sec o dummies, and 18
ime dummies.
Wo ks Councils, So ing and Pe o mance / 455

P oduc i i y. Table 4 p esen s ou labo p oduc i i y OLS eg essions. The
ocus is on he e ec o council exis ence and how wo ke quali y shapes he
e ec . The fi s column is no con olling o wo ke quali y and shows ha
TABLE 4
LABOR PRODUCTIVITY, OLS REGRESSIONS
(1) (2) (3) (4) (5) (6)
Wo ks council 0.160*** 0.145*** 0.128*** 0.090*** 0.121*** 0.075***
(0.023) (0.023) (0.022) (0.027) (0.022) (0.027)
Skilled employees 0.367*** 0.278*** 0.370*** 0.269*** 0.281***
(0.029) (0.028) (0.029) (0.029) (0.028)
Wo ke quali y 0.104*** 0.100*** 0.104***
(0.009) (0.009) (0.009)
Collec i e ba gaining −0.002 −0.014 −0.015 −0.039** −0.015 −0.039**
(0.015) (0.014) (0.014) (0.017) (0.014) (0.017)
Wo ks council ×collec i e
ba gaining
0.090*** 0.088***
(0.031) (0.030)
Wo ks council ×wo ke
quali y
0.037**
(0.017)
Log(capi al in ensi y) 0.099*** 0.095*** 0.092*** 0.094*** 0.091*** 0.091***
(0.007) (0.007) (0.007) (0.007) (0.007) (0.007)
Expo e 0.120*** 0.122*** 0.108*** 0.121*** 0.109*** 0.107***
(0.016) (0.016) (0.015) (0.016) (0.015) (0.015)
Single plan −0.137*** −0.123*** −0.109*** −0.121*** −0.109*** −0.107***
(0.016) (0.016) (0.015) (0.016) (0.015) (0.015)
Technical s a e =good −0.056*** −0.049*** −0.046*** −0.049*** −0.046*** −0.046***
(0.013) (0.013) (0.013) (0.013) (0.013) (0.013)
Technical s a e = ai −0.108*** −0.089*** −0.081*** −0.090*** −0.081*** −0.082***
(0.017) (0.017) (0.016) (0.017) (0.016) (0.016)
Technical s a e =poo −0.152*** −0.126*** −0.116*** −0.127*** −0.117*** −0.117***
(0.030) (0.029) (0.029) (0.029) (0.029) (0.029)
Pa - ime employees −0.983*** −0.894*** −0.892*** −0.893*** −0.896*** −0.892***
(0.052) (0.049) (0.049) (0.049) (0.050) (0.049)
App en ices −0.912*** −0.806*** −0.785*** −0.789*** −0.794*** −0.769***
(0.106) (0.105) (0.104) (0.106) (0.104) (0.104)
Female employees −0.113** −0.060 −0.002 −0.058 0.002 0.000
(0.045) (0.044) (0.044) (0.044) (0.044) (0.044)
Chu ning a e −0.035 −0.022 −0.016 −0.022 −0.016 −0.016
(0.050) (0.040) (0.037) (0.040) (0.038) (0.037)
Cons an 10.036*** 9.739*** 9.807*** 9.755*** 9.814*** 9.823***
(0.370) (0.352) (0.299) (0.347) (0.301) (0.293)
R
2
0.385 0.400 0.416 0.400 0.416 0.416
N22,576 22,576 22,576 22,576 22,576 22,576
No es: LIAB c oss-sec ional model, 1998–2016, Wes Ge many, OLS sample. ***/**/*deno es s a is ical significance a
he 1%/5%/10% le el. The dependen a iable is he loga i hm o he alue added di ided by he numbe o employees.
Wo ke quali y is he mean o he AKM wo ke e ec s (α
i
) a he plan le el (as desc ibed in Sec ion “Da a and Empi i-
cal S a egy”) s anda dized wi h a mean o ze o and a s anda d de ia ion o one. Repo ed numbe s a e coe ficien s om
OLS eg essions wi h s anda d e o s clus e ed a he plan le el in pa en heses. Fu he co a ia es included in all specifi-
ca ions a e 7 plan size dummies, 8 ede al s a e dummies, 37 wo-digi sec o dummies and 18 ime dummies.
456 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
council plan s a e ce e is pa ibus 16 pe cen mo e p oduc i e. Adding he sha e
o skilled wo ke s in column (2) yields a posi i e impac o skill on p oduc i -
i y and a educ ion o he council coe ficien om 0.160 o 0.145. Wi h his
esul , we a e in he same ange o magni ude as o he ecen s udies (compa e,
e.g., Ji jahn and Muelle 2014; Muelle 2015). Including AKM wo ke e ec s
(column 3) yields a s ong posi i e impac o hem on p oduc i i y and a u -
he educ ion o he council e ec om 0.145 o 0.128.
25
This leads o wo
conclusions: fi s , p ope ly con olling o wo ke quali y educes he council
e ec by abou 12 pe cen bu he e is s ill a subs an ial p oduc i i y e ec le ;
and second, AKM wo ke e ec s a e s ongly ela ed o p oduc i i y e en i
he pe cen age o skilled jobs is con olled o .
26
Coe ficien s o co a ia es no a he cen e s age o ou analysis show no
su p ises; ha is, plan s ha expo , belong o mul i-b anch fi ms, use mo e
capi al pe wo ke and mo e up- o-da e equipmen , and employ less app en-
ices and pa - ime wo ke s ha e ce e is pa ibus highe labo p oduc i i y.
27
In
column (4), we confi m he s ongly posi i e in e ac ion e ec be ween coun-
cils and collec i e ag eemen s. Column (5) shows ha in e ac ing wo ke qual-
i y and council s a us yields a significan and posi i e coe ficien , which means
ha he e ec o wo ke quali y on p oduc i i y is by one hi d la ge in coun-
cil plan s. This leads o he conclusion ha while council plan s do employ
be e wo ke s as documen ed in Tables 1 and 2, hey a e also making be e
use o hem. Column (6) finally documen s ha he in e ac ion e m be ween
wo ks council p esence and collec i e ag eemen s is no shaped by con olling
o AKM e ec s.
Figu e 2 displays he e en s udy es ima es o labo p oduc i i y. Confi m-
ing Muelle and S egmaie (2017), Panel A shows ha plan s in oducing a
wo ks council expe ience a down u n in p oduc i i y be o e he in oduc ion
and inc easing p oduc i i y as he council g ows olde . This is in line wi h he
findings in Ji jahn (2009), K a and Lang (2008) and Moh enweise e al.
(2012) who show ha wo ks council in oduc ions a e mo e likely when he
plan is unde economic dis ess. The nega i e p e- end o council adop e s
implies ha he posi i e e ec s we measu e a e council in oduc ion migh
25
Albei being e y p ecisely es ima ed, he di e ence in council coe ficien s is no s a is ically signifi-
can .
26
No e ha he ew p e ious s udies on he economic e ec s o wo ks councils ha employ plan fixed
e ec s (e.g., Addison e al. 2004) implici ly also con ol o unobse ed wo ke he e ogenei y. This, how-
e e , comes a he cos o only being able o use wi hin-fi m a ia ion in wo ks council s a us (and in all
o he a iables, oo) and o no being able o ac ually pin down he e ec o wo ke so ing and i s in e ac-
ion wi h o he a iables.
27
We do no con ol o he owne ship s uc u e in ou eg essions as his would educe ou sample size
wi hou changing any esul s.
Wo ks Councils, So ing and Pe o mance / 457
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 6 o 8 ≥ 9
ela i e ime in e als
baseline wi hou con ol a iables
90% con idence in e als
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 6 o 8 ≥ 9
ela i e ime in e als
baseline condi ional on wo ke
q
uali
y
FIGURE 2
LABOR PRODUCTIVITY,EVENT STUDY.
458 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
e en unde s a e he causal p oduc i i y e ec o wo ks councils. When adding
wo ke quali y o he e en s udy (Panel B), we find ha he g ow h in p oduc-
i i y is no d i en by he upg ade in wo ke quali y. This is no su p ising
because a la ge po ion o he wo ke quali y ad an age o council plan s
al eady exis ed p io o council in oduc ion (see Table 1) and, hus, is cap-
u ed by he fixed e ec . We hus suppo Muelle and S egmaie (2017) in
hei conclusion ha he p oduc i i y inc ease is likely o be a genuine council
e ec .
Wages. Table 5 shows ou OLS wage es ima es. The coe ficien s o he
con ol a iables mos ly ha e he same sign as in he p oduc i i y eg essions,
unde lining he close link be ween p oduc i i y and wages. Wi hou con olling
o skill equi emen s and wo ke quali y, wo ks councils a e ce e is pa ibus
associa ed wi h 12 pe cen highe wages (column 1), which d ops o 10 pe -
cen when he sha e o skilled wo ke s is added (column 2). Ou es ima es a e
smalle han, o example, hose in Addison e al. (2001) who epo ed abou
15 pe cen highe wages. Adding AKM wo ke e ec s educes he council
wage p emium u he o abou 8 pe cen (column 3). The ela i ely mild
educ ion o he council coe ficien shows ha he council p emium is no ully
explained by he council plan s’be e wo ke s. I a he suppo s he no ion
ha ac o s such as he wo ke s’ba gaining powe d i e he council p emium
(Hi sch and Muelle 2020). Ou esul s show ha one s anda d de ia ion
inc ease in AKM wo ke e ec s is associa ed wi h a wage inc ease o 11 pe -
cen (column 3), condi ional on he sha e o skilled jobs.
No es: LIAB c oss-sec ional model, 1998–2016, Wes Ge many, e en s udy sample (141,098
plan -yea -coho obse a ions). Wo ks council in oduc ions be ween 1998 and 2016. This
figu e shows he mean ou come o loga i hm o alue added di ided by he numbe o
employees ela i e o he p e-in oduc ion pe iod o he wo ks council and ne o he e olu ion
in he con ol g oup. Wo ke quali y is measu ed as he mean o he AKM wo ke e ec s (α
i
)
a he plan le el (as desc ibed in Sec ion “Da a and Empi ical S a egy”) s anda dized wi h a
mean o ze o and a s anda d de ia ion o one. The e en (yea =0 o 2) is he in oduc ion
pe iod o he wo ks council and ela i e ime (in yea s) is depic ed a he ho izon al axis. As
specified in equa ion (3), he eg ession includes con ols o collec i e wage ag eemen
p esence, capi al in ensi y, expo s a us, single-plan s a us, he s a e o echnical machine y,
he sha e o skilled employees, pa - ime wo ke s, app en ices and women o all employees, 7
plan size dummies, and plan -coho fixed e ec s. Fo panel A: No con ols o he han
specified in equa ion (3) a e used. Fo panel B: holding cons an wo ke quali y. The 90%
confidence in e als a e shown using s anda d e o s clus e ed a he plan le el [Colo figu e
can be iewed a wileyonlinelib a y.com]
Wo ks Councils, So ing and Pe o mance / 459
in alida ing he gene al esul o posi i e council e ec s as documen ed in he
moun ing li e a u e on wo ks councils.
Finally, we documen ed a posi i e link be ween council exis ence and plan
p ofi abili y e en a e con olling o wo ke quali y. Councils seem o make
su e ha he p oduc i i y gains associa ed wi h hem a e spli be ween labo
and capi al. In combina ion, ou fixed-e ec s e en s udy esul s show ha
plan s expe ience u bulen imes be o e council in oduc ion wi h s ong wage
g ow h and a subs an ial p oduc i i y decline ha sha ply educes p e-
in oduc ion p ofi s. A e council in oduc ion, wage g ow h fla ens and p o-
duc i i y g ow h se s in, which allows council plan s o sus ain long- un p o -
i abili y wi hin a high-wage high-p oduc i i y s a egy.
We conclude ha councils con ibu e o p oduc i i y, wage, and p ofi
inequali y ac oss plan s, fi s , by a ac ing and sus aining high-wage high-
p oduc i i y wo ke s and, second, by a genuine council e ec on fi m pe o -
mance. We show s ong posi i e p oduc i i y con ibu ions o high-wage wo k-
e s ha a e e en s onge when wo ks councils a e p esen . This lends suppo
o he no ion ha wo ke quali y and wo ke pa icipa ion, as a o m o high-
pe o mance managemen p ac ices, a e complemen s. We conclude ha so ing
o high-quali y wo ke s o wo ks council plan s can imp o e alloca i e e fi-
ciency and agg ega e p oduc i i y.
No es: LIAB c oss-sec ional model, 1998–2016, Wes Ge many, e en s udy sample (141,098
plan -yea -coho obse a ions). Wo ks council in oduc ions be ween 1998 and 2016. This
figu e shows he mean ou come o loga i hm o he alue added minus labo cos s di ided by
he numbe o employees ela i e o he p e-in oduc ion pe iod o he wo ks council and ne
o he e olu ion in he con ol g oup. Wo ke quali y is measu ed as he mean o he AKM
wo ke e ec s (α
i
) a he plan le el (as desc ibed in Sec ion “Da a and Empi ical S a egy”)
s anda dized wi h a mean o ze o and a s anda d de ia ion o one. The e en (yea =0 o2)
is he in oduc ion pe iod o he wo ks council and ela i e ime (in yea s) is depic ed a he
ho izon al axis. As specified in equa ion (3), he eg ession includes con ols o collec i e
wage ag eemen p esence, capi al in ensi y, expo s a us, single-plan s a us, he s a e o
echnical machine y, he sha e o skilled employees, pa - ime wo ke s, app en ices and
women o all employees, 7 plan size dummies, and plan -coho fixed e ec s. Fo panel A:
No con ols o he han specified in equa ion (3) a e used. Fo panel B: holding cons an
wo ke quali y. The 90% confidence in e als a e shown using s anda d e o s clus e ed a he
plan le el [Colo figu e can be iewed a wileyonlinelib a y.com]
466 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER

REFERENCES
Abowd, J. M., F. K ama z, and D. N. Ma golis. 1999. “High Wage Wo ke s and High Wage Fi ms.”Econo-
me ica 67(2): 251–333.
Adam, J. B. 2019. “Volun a y Qui s: Do Wo ks Councils Ma e ? An Analysis o he Re o m o he Ge man
Wo ks Cons i u ion Ac 2001.”Jah b¨
uche F¨
u Na ional¨
okonomie Und S a is ik 239(1): 67–109.
Addison, J. T. 2009. The Economics o Code e mina ion: Lessons om he Ge man Expe ience. New Yo k,
NY: Palg a e Macmillan.
Addison, J. T., A. B yson, P. Teixei a, A. Pahnke, and L. Bellmann. 2013. “The Ex en o Collec i e Ba -
gaining and Wo kplace Rep esen a ion: T ansi ions be ween S a es and hei De e minan s. A Compa -
a i e Analysis o Ge many and G ea B i ain.”Sco ish Jou nal o Poli ical Economy 60: 182–209.
Addison, J. T., C. Schnabel, and J. Wagne . 2001. “Wo ks Councils in Ge many: Thei E ec s on Es ablish-
men Pe o mance.”Ox o d Economic Pape s 53(4): 659–94.
Addison, J. T., C. Schnabel, and J. Wagne . 2004. “The Cou se o Resea ch in o he Economic Conse-
quences o Ge man Wo ks Councils.”B i ish Jou nal o Indus ial Rela ions 42(2): 255–81.
Addison, J. T., P. Teixei a, and T. Zwick. 2010. “Ge man Wo ks Councils and he Ana omy o Wages.”ILR
Re iew 63(2): 247–70.
Addison, J. T., and J. Wagne . 1997. “De Einfluss on Be iebs ¨
a en au P ofi abili ¨
a und P oduk inno a io-
nen: neue Un e suchungse gebnisse au den Basis on Be iebsda en.”Jah b¨
uche F¨
u Na io-
nal¨
okonomie Und S a is ik 216(1): 1–20.
And ews, M. J., L. Gill, T. Schank, and R. Upwa d. 2008. “High Wage Wo ke s and Low Wage Fi ms:
Nega i e Asso a i e Ma ching o Limi ed Mobili y Bias?”Jou nal o he Royal S a is ical Socie y:
Se ies A (S a is ics in Socie y) 171(3): 673–97.
And ews, M. J., L. Gill, T. Schank, and R. Upwa d. 2012. “High Wage Wo ke s Ma ch wi h High Wage
Fi ms: Clea E idence o he E ec s o Limi ed Mobili y Bias.”Economics Le e s 117(3): 824–7.
Beh ens, M. 2009. “S ill Ma ied a e All These Yea s? Union O ganizing and he Role o Wo ks Councils
in Ge man Indus ial Rela ions.”ILR Re iew 62(3): 275–93.
Bellmann, L., B. Lochne , S. Se h, and S. Wol e 2020. “AKM e ec s o Ge man labou ma ke da a.”
FDZ-Me hoden epo , 01/2020, Nu embe g.
Bende , S., N. Bloom, D. Ca d, J. Van Reenen, and S. Wol e . 2018. “Managemen P ac ices, Wo k o ce
Selec ion, and P oduc i i y.”Jou nal o Labo Economics 36(S1): S371–409.
Bonhomme, S., T. Lamadon, and E. Man esa. 2019. “A Dis ibu ional F amewo k o Ma ched Employe
Employee Da a.”Econome ica 87(3): 699–739.
B ¨
andle, T. 2017. “Flexible Collec i e Ba gaining Ag eemen s: S ill a Mode a ing E ec on Wo ks Council
Beha iou ?”Manage ial and Decision Economics 38(8): 1137–53.
Ca d, D., J. Heining, and P. Kline. 2013. “Wo kplace He e ogenei y and he Rise o Wes Ge man Wage
Inequali y.”The Qua e ly Jou nal o Economics 128(3): 967–1015.
Dobbelae e, S., B. Hi sch, S. Muelle , and G. Neuschae e . 2020. “O ganised Labou , Labou Ma ke
Impe ec ions, and Employe Wage P emia.”IZA DP No. 13909, No embe 2020.
Ellgu h, P., and S. Kohau . 2016. “Ta i bindung und be iebliche In e essen e e ung: E gebnisse aus dem
IAB-Be iebspanel 2015.”WSI-Mi eilungen 69(4): 283–91.
Ellgu h, P., S. Kohau , and I. M¨
olle . 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.
Fi po, S., N. M. Fo in, and T. Lemieux. 2009. “Uncondi ional Quan ile Reg essions.”Econome ica 77(3):
953–73.
Fi z oy, F. R., and K. K a . 1987. “E ciency and In e nal O ganiza ion: Wo ks Councils in Wes Ge man
Fi ms.”Economica 54(216): 493–504.
F eeman, R. 1976. “Indi idual Mobili y and Union Voice in he Labo Ma ke .”Ame ican Economic Re iew
66(2): 361–8.
F eeman, R. B., and E. P. Lazea . 1995. “An Economic Analysis o Wo ks Councils.”In Wo ks Councils:
Consul a ion, Rep esen a ion, and Coope a ion in Indus ial Rela ions, edi ed by J. Roge s and W.
S eeck, pp. 27–52. Chicago, IL: Uni e si y o Chicago P ess.
Wo ks Councils, So ing and Pe o mance / 467
F ick, B. 1996. “Co-de e mina ion and Pe sonnel Tu no e : The Ge man Expe ience.”Labou 10(2): 407–
30.
F ick, B., and I. M¨
olle . 2003. “Manda ed Wo ks Councils and Fi m Pe o mance: Labo P oduc i i y and
Pe sonnel Tu no e in Ge man Es ablishmen s.”Schmolle s Jah buch (Jou nal o Applied Social
Science S udies) 123(3): 423–54.
G und, C., J. Ma in, and A. Schmi . 2016. “Wo ks Councils, Qui s and Dismissals.”Ge man Jou nal o
Human Resou ce Managemen 30(1): 53–75.
Gue zgen, N. 2009. “Ren –sha ing and Collec i e Ba gaining Co e age: E idence om Linked Employe -
Employee Da a.”Scandina ian Jou nal o Economics 111(2): 323–49.
Heining, J., W. Klos e hube , and S. Se h. 2014. “An O e iew on he Linked Employe -Employee Da a o
he Ins i u e o Employmen Resea ch (IAB).”Schmolle s Jah buch 134(1): 141–8.
Hijzen, A., R. Upwa d, and P. W. W igh . 2010. “The Income Losses o Displaced Wo ke s.”Jou nal o
Human Resou ces 45(1): 243–69.
Hi sch, B., and S. Muelle . 2020. “Fi m Wage P emia, Indus ial Rela ions, and Ren Sha ing in Ge many.”
ILR Re iew 73(5): 1119–46. h ps://doi.o g/10.1177/0019793920917105.
Hi sch, B., T. Schank, and C. Schnabel. 2010. “Wo ks Councils and Sepa a ions: Voice, Monopoly, and
Insu ance E ec s.”Indus ial Rela ions 49(4): 566–92.
Hi schman, Albe O. 1970. “Exi , Voice and Loyal y: Responses o Decline in Fi ms, O ganiza ions, and
S a es.”The Jou nal o Finance 25(5): 1194.
Huble , O., and U. Ji jahn. 2003. “Wo ks Councils and Collec i e Ba gaining in Ge many: The Impac on
P oduc i i y and Wages.”Sco ish Jou nal o Poli ical Economy 50(4): 471–91.
Janssen, S., and J. Moh enweise . 2018. “The shel li e o incumben wo ke s du ing accele a ing echnologi-
cal change. E idence om a aining egula ion e o m.”IZA discussion pape , 11312, Bonn.
Jensen, M. C., and W. H. Meckling. 1979. “Righ s and P oduc ion Func ions: An Applica ion o Labo -
Managed Fi ms and Code e mina ion.”The Jou nal o Business 52(4): 469–506.
Ji jahn, U. 2009. “The In oduc ion o Wo ks Councils in Ge man Es ablishmen s –Ren Seeking o Ren
P o ec ion?”B i ish Jou nal o Indus ial Rela ions 47(3): 521–45.
Ji jahn, U., and K. K a . 2011. “Do Spillo e s S imula e Inc emen al o D as ic P oduc Inno a ions? E i-
dence om Ge man Es ablishmen Da a.”Ox o d Bulle in o Economics and S a is ics 73(4): 209–538.
Ji jahn, U., and J. Moh enweise . 2016. “Owne -manage s and he Failu e o Newly Adop ed Wo ks Coun-
cils.”B i ish Jou nal o Indus ial Rela ions 54(4): 815–45.
Ji jahn, U., J. Moh enweise , and U. Backes-Gellne . 2011. “Wo ks Councils and Lea ning: On he Dynamic
Dimension o Code e mina ion.”Kyklos 64(3): 427–47.
Ji jahn, U., and S. Muelle . 2014. “Non-union Wo ke Rep esen a ion, Fo eign Owne s, and he Pe o mance
o Es ablishmen s.”Ox o d Economic Pape s 66(1): 140–63.
Ji jahn, U., and S. C. Smi h. 2018. “Nonunion Employee Rep esen a ion: Theo y and he Ge man Expe i-
ence wi h Manda ed Wo ks Councils.”Annals o Public and Coope a i e Economics 89(1): 201–33.
K a , K., and J. Lang. 2008. “The Causes and Consequences o Adop ing a Wo ks Council.”Jah b¨
uche
F¨
u Na ional¨
okonomie Und S a is ik 228(5+6): 512–32.
Lazea , E. P. 2000. “Pe o mance Pay and P oduc i i y.”Ame ican Economic Re iew 90(5): 1346–61.
Lopes de Melo, R. 2018. “Fi m Wage Di e en ials and Labo Ma ke So ing: Reconciling Theo y and E i-
dence.”Jou nal o Poli ical Economy 126(1): 313–46.
Moh enweise , J., P. Ma ginson, and U. Backes-Gellne . 2012. “Wha T igge s he Es ablishmen o a Wo ks
Council?”Economic and Indus ial Democ acy 33(2): 295–316.
Muelle , S. 2008. “Capi al S ock App oxima ion using Fi m Le el Panel Da a.”Jah b¨
uche F¨
u Na io-
nal¨
okonomie Und S a is ik 228(4): 357–71.
Muelle , S.. 2010. “Capi al s ock app oxima ion wi h he pe pe ual in en o y me hod *s a a code o he
IAB es ablishmen panel.”FDZ-Me hoden epo , 02/2010, Nu embe g.
468 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
Muelle , S. 2011. “Wo ks Councils and Fi m P ofi s Re isi ed.”B i ish Jou nal o Indus ial Rela ions 49
(s1): s27–43.
Muelle , S. 2012. “Wo ks Councils and Es ablishmen P oduc i i y.”ILR Re iew 65(4): 880–98.
Muelle , S. 2015. “Wo ks Councils and Labou P oduc i i y: Looking Beyond he Mean.”B i ish Jou nal o
Indus ial Rela ions 53(2): 308–25.
Muelle , S. 2017. “Capi al s ock app oxima ion wi h he pe pe ual in en o y me hod: An upda e.”FDZ-
Me hoden epo 05/2017, Nu embe g.
Muelle , S., and J. S egmaie . 2017. “The Dynamic E ec s o Wo ks Councils on Labou P oduc i i y: Fi s
E idence om Panel Da a.”B i ish Jou nal o Indus ial Rela ions 55(2): 372–95.
M¨
ulle , S., and J. S egmaie . 2020. “Why is The e Resis ance o Wo ks Councils in Ge many? An Economic
Pe spec i e.”Economic and Indus ial Democ acy 41(3): 540–61.
Obe fich ne , M. 2019. “Wo ks Council In oduc ions in Ge many: Do They Reflec Wo ke s’Voice?”Eco-
nomic and Indus ial Democ acy 40(2): 301–25.
Obe fich ne , M., and C. Schnabel. 2019. “The Ge man Model o Indus ial Rela ions: (Whe e) does i S ill
Exis ?”Jah b¨
uche F¨
u Na ional¨
okonomie Und S a is ik 239(1): 5–37.
P ei e , C. 2011. “Wo ks Councils, Union Ba gaining and Qui s in Ge man Fi ms.”Economic and Indus ial
Democ acy 32(2): 243–60.
Schmid lein, L., S. Se h, and M. Umkeh e . 2019. “Linked employe -Employee Da a om he IAB: LIAB
C oss-Sec ional Model 2 (LIAB QM2) 1993-2017.”FDZ-Me hoden epo , 06/2019, Nu embe g.
Schnabel, C., and J. Wagne . 1994. “Indus ial Rela ions and T ade Union E ec s on Inno a ion in Ge -
many.”Labou 8(3): 489–504.
S egmaie , J. 2012. “E ec s o Wo ks Councils on Fi m-P o ided Fu he T aining in Ge many.”B i ish
Jou nal o Indus ial Rela ions 50(4): 667–89.
Sy e son, C. 2011. “Wha De e mines P oduc i i y?”Jou nal o Economic Li e a u e 49(2): 326–65.
Wagne , J., T. Schank, C. Schnabel, and J. T. Addison. 2006. “Wo ks Councils, Labo P oduc i i y and Plan
He e ogenei y: Fi s E idence om Quan ile Reg essions.”Jou nal o Economics and S a is ics
(Jah b¨
uche F¨
u Na ional¨
okonomie Und S a is ik) 226(5): 505–18.
Wol , E., and T. Zwick. 2002. “P oduk i i ¨
a swi kung on Mi a bei e be eiligung: De Einfluss on unbeo-
bach e e He e ogeni ¨
a .”Mi eilungen Aus De A bei sma k - Und Be u s o schung 35(1): 123–32.
Wo ks Councils, So ing and Pe o mance / 469
APPENDIX A
Defini ions o Va iables
TABLE A1
DEFINITIONS OF VARIABLES
Va iable Defini ion
Log(labo p oduc i i y) Loga i hm o he alue added pe wo ke
Log(wage bill pe wo ke ) Loga i hm o he wage bill pe wo ke
Log(p ofi pe wo ke ) Loga i hm o he alue added ne o wage cos s (including employe s’
social secu i y con ibu ions) pe wo ke
Log(employmen ) Loga i hm o he numbe o wo ke s
Log(capi al in ensi y) Loga i hm o he capi al s ock pe wo ke
Wo ks council =1 i a wo ks council is p esen ,
=0 i no wo ks council is p esen
Collec i e ba gaining =1 i collec i e ba gaining is p esen ,
=0 i no collec i e ba gaining is p esen
Wo ke quali y Mean o he AKM wo ke e ec s (α
i
) a he plan le el (as desc ibed
in Sec ion “Da a and Empi ical S a egy”) s anda dized wi h a mean
o 0 and a s anda d de ia ion o 1
Skilled employees as Sha e o wo ke s who ha e a oca ional qualifica ion,
sha e o all employees ele an p o essional expe ience, o an uni e si y deg ee
Pa - ime employees as Sha e o pa - ime wo ke s
sha e o all employees
App en ices as Sha e o wo ke s who a e doing hei oca ional aining unde he
oca ional aining law o he Handic a s Regula ion Ac and o he
aining s ipula ions o all wo ke s
sha e o all employees
Female employees Sha e o women
sha e o all employees
Chu ning a e Measu e o employmen s abili y. Wo ke flow a e minus he absolu e
alue o he ne a e o employmen change.
Expo e =1 i plan makes e enue ab oad,
=0 i plan does no make e enue ab oad
Single plan =1 i he plan is an independen company o an independen
o ganiza ion wi hou any o he places o business, =0 i plan does
ha e o he /belongs o o he b anches
Technical s a e o machine y Assessmen o he o e all s a e o he echnical s a e o he plan and
machine y compa ed wi h o he plan s in he same indus y. Scale
om 1 o 5.
excellen =1
good =2
ai =3
poo =4 and 5
A e age wo ke age A e age age o all employees
Uni e si y g adua es as Sha e o uni e si y g adua es
sha e o all employees
No es: Linked Employe –Employee Da a o he IAB (LIAB), c oss-sec ional model.
470 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
APPENDIX B
Robus ness Checks
Selec i e panel a i ion. In his sec ion, we p esen obus ness checks o
he esul s in Sec ion 4 ega ding selec i e panel a i ion. Panel a i ion is a
ea u e o mos panel da a se s and may also be an issue in ou s. So a , we
use in o ma ion on he obse ed su ey yea s o 2012, 2014, and 2016. I
being obse ed in one o he h ee yea s is mo e likely o success ul council
plan s, we o e sample success ul council plan s because unsuccess ul council
plan s d opped ou (su i o ship bias) ea lie .
To add ess selec i e panel a i ion, we conduc an e en s udy in which we
only include plan s, o which we di ec ly obse e council in oduc ion in ou
da a. This means ha his sample includes all young wo ks councils, ega dless
o he quali y o he council plan o he council i sel and ega dless whe he
he plan su i es un il he yea s whe e council age is su eyed (i.e., 2012,
2014, 2016). The esul s a e depic ed in Figu es B1-B4 and show o each ou -
come he same pa e s as in ou baseline esul s p esen ed in Sec ion 4 ha
elied on he council age su ey ques ion. We include wo pos -e en ime
dummies ins ead o h ee, because highe wo ks council age ca ego ies a e
poo ly filled.
Wo ks Councils, So ing and Pe o mance / 471

-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
baseline wi hou con ol a iables
90% con idence in e als
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
baseline condi ional on wo ke age
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
baseline condi ional on uni e si y deg ee
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
baseline condi ional on age and educa ion
FIGURE B1
WORKER QUALITY,EVENT STUDY (YOUNG WORKS COUNCILS).
No es:. LIAB c oss-sec ional model, 1998–2016, Wes Ge many, e en s udy sample (140,782
plan -yea -coho obse a ions). Wo ks council in oduc ions be ween 1998 and 2016. This
figu e shows he mean ou come o wo ke quali y ela i e o he p e-in oduc ion pe iod o he
wo ks council and ne o he e olu ion in he con ol g oup. Wo ke quali y is measu ed as he
mean o he AKM wo ke e ec s (α
i
) a he plan le el (as desc ibed in Sec ion “Da a and
Empi ical S a egy”) s anda dized wi h a mean o ze o and a s anda d de ia ion o one. The
ea men g oup o his sample includes only wo ks council in oduc ions, which a e
de e mined using he panel s uc u e o he da a only. The e en (yea =0 o 2) is he
in oduc ion pe iod o he wo ks council, and ela i e ime (in yea s) is depic ed a he
ho izon al axis. As specified in equa ion (3), he eg ession includes con ols o collec i e
wage ag eemen p esence, capi al in ensi y, expo s a us, single-plan s a us, he s a e o
echnical machine y, he sha e o skilled employees, pa - ime wo ke s, app en ices and
women o all employees, 7 plan size dummies, and plan -coho fixed e ec s. Fo panel A:
No con ols o he han specified in equa ion (3) a e used. Fo panel B: holding cons an
wo ke age a he plan le el. Fo Panel C: holding cons an he combined sha e o uni e si y
g adua es. Fo Panel D: holding cons an age and he sha e o uni e si y g adua es. The 90%
confidence in e als a e shown using s anda d e o s clus e ed a he plan le el [Colo figu e
can be iewed a wileyonlinelib a y.com]
472 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
baseline wi hou con ol a iables
90% con idence in e als
-0.20
-0.10
0.00
0.10
0.20
0.30
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
base
lin
eco
n
d
i
i
o
n
a
l
o
n w
o
k
e
qua
li
y
FIGURE B2
LABOR PRODUCTIVITY,EVENT STUDY (YOUNG WORKS COUNCILS).
Wo ks Councils, So ing and Pe o mance / 473
No es. LIAB c oss-sec ional model, 1998–2016, Wes Ge many, e en s udy sample (140,782
plan -yea -coho obse a ions). Wo ks council in oduc ions be ween 1998 and 2016. This
figu e shows he mean ou come o alue added di ided by he numbe o employees ela i e
o he p e-in oduc ion pe iod o he wo ks council and ne o he e olu ion in he con ol
g oup. Wo ke quali y is measu ed as he mean o he AKM wo ke e ec s (α
i
) a he plan
le el (as desc ibed in Sec ion “Da a and Empi ical S a egy”) s anda dized wi h a mean o
ze o and a s anda d de ia ion o one. The ea men g oup o his sample includes only wo ks
council in oduc ions, which a e de e mined using he panel s uc u e o he da a only. The
e en (yea =0 o 2) is he in oduc ion pe iod o he wo ks council, and ela i e ime (in
yea s) is depic ed a he ho izon al axis. As specified in equa ion (3), he eg ession includes
con ols o collec i e wage ag eemen p esence, capi al in ensi y, expo s a us, single-plan
s a us, he s a e o echnical machine y, he sha e o skilled employees, pa - ime wo ke s,
app en ices and women o all employees, 7 plan size dummies, and plan -coho fixed e ec s.
Fo panel A: No con ols o he han specified in equa ion (3) a e used. Fo panel B: holding
cons an wo ke quali y. The 90% confidence in e als a e shown using s anda d e o s
clus e ed a he plan le el [Colo figu e can be iewed a wileyonlinelib a y.com]
474 / STEFFEN MUELLER AND GEORG NEUSCHAEFFER
-0.20
-0.10
0.00
0.10
0.20
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
baseline wi hou con ol a iables
90% con idence in e als
-0.20
-0.10
0.00
0.10
0.20
e en ime coe icien
≤ -4 -3 o -1 0 o 2 3 o 5 ≥ 6
yea s since in oduc ion
baseline condi ional on wo ke
q
uali
y
FIGURE B3
WAGES,EVENT STUDY (YOUNG WORKS COUNCILS).
Wo ks Councils, So ing and Pe o mance / 475