Family firms and their role in the fall of the labor share and the rise of corporate saving in Germany

Beh inge , Jan; an T eeck, Till; Vic o , Vincen
Wo king Pape
Family i ms and hei ole in he all o he labo sha e
and he ise o co po a e sa ing in Ge many
FMM Wo king Pape , No. 115
P o ided in Coope a ion wi h:
Mac oeconomic Policy Ins i u e (IMK) a he Hans Boeckle Founda ion
Sugges ed Ci a ion: Beh inge , Jan; an T eeck, Till; Vic o , Vincen (2025) : Family i ms and hei ole
in he all o he labo sha e and he ise o co po a e sa ing in Ge many, FMM Wo king Pape , No.
115, Hans-Böckle -S i ung, Mac oeconomic Policy Ins i u e (IMK), Fo um o Mac oeconomics and
Mac oeconomic Policies (FMM), Düsseldo
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/324480
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/legalcode.de
FMM WORKING PAPER
No. 115 • Feb ua y 2025 • Hans-Böckle -S i ung
FAMILY FIRMS AND THEIR ROLE
IN THE FALL OF THE
LABOR SHARE
AND
THE RISE OF CORPORATE
SAVING IN GERMANY
Jan Beh inge
1
, Till an T eeck
2
, Vincen Vic o
3
ABSTRACT
This pape in es iga es he ole o amily i ms in he all o he labo sha e and ise in
co po a e sa ing in Ge many om 1993 o 2019. Combining a new Family Owne ship and
Go e nance (FOG) da abase wi h inancial da a, we analyze 929 publicly lis ed i ms. Ou
indings show ha i m-le el labo sha e declines a e widesp ead in Ge many, con as ing
wi h indings om he U.S. ha link his end o a ew as -g owing supe s a i ms. Family
i ms, pa icula ly in manu ac u ing, expe ienced sha pe dec eases in he labo sha e and
s onge inc eases in co po a e sa ing compa ed o non- amily i ms. The le el o amily
in ol emen in Ge many’s wo- ie boa d sys em (managemen and supe iso y boa d)
u he a ec s hese ou comes. Despi e paying lowe wages, we ind no e idence ha amily
i ms p o ide g ea e employmen s abili y. Ou esul s challenge global gene aliza ions
abou he d i e s o he labo sha e and co po a e sa ing, while emphasizing he
mac oeconomic ele ance o amily i ms, especially in Ge many’s co po a e sec o .
—————————
1 Mac oeconomic Policy Ins i u e (IMK). Email: jan-beh [email p o ec ed]
2 Ins i u e o Socio-Economics, Uni e si y o Duisbu g-Essen. Email: ill. an [email protected]
3 Ins i u e o Socio-Economics, Uni e si y o Duisbu g-Essen. Email: incen . ic o @s ud.uni-due.de
Family i ms and hei ole in he all o he labo sha e and
he ise o co po a e sa ing in Ge many*
Jan Beh inge
Mac oeconomic Policy Ins i u e (IMK)
Till an T eeck
Uni e si y o Duisbu g-Essen
Vincen Vic o
Uni e si y o Duisbu g-Essen
Feb ua y 20, 2025
Abs ac
This pape in es iga es he ole o amily i ms in he all o he labo sha e and ise in
co po a e sa ing in Ge many om 1993 o 2019. Combining a new Family Owne ship and
Go e nance (FOG) da abase wi h inancial da a, we analyze 929 publicly lis ed i ms. Ou
indings show ha i m-le el labo sha e declines a e widesp ead in Ge many, con as ing wi h
indings om he U.S. ha link his end o a ew as -g owing supe s a i ms. Family i ms,
pa icula ly in manu ac u ing, expe ienced sha pe dec eases in he labo sha e and s onge
inc eases in co po a e sa ing compa ed o non- amily i ms. The le el o amily in ol emen
in Ge many’s wo- ie boa d sys em (managemen and supe iso y boa d) u he a ec s hese
ou comes. Despi e paying lowe wages, we ind no e idence ha amily i ms p o ide g ea e
employmen s abili y. Ou esul s challenge global gene aliza ions abou he d i e s o he
labo sha e and co po a e sa ing, while emphasizing he mac oeconomic ele ance o amily
i ms, especially in Ge many’s co po a e sec o .
Keywo ds: Labo sha e, co po a e sa ing, amily i ms
JEL Classi ica ions: D22, D33, G32
*Jan Beh inge , Mac oeconomic Policy Ins i u e (IMK), Geo g-Glock-S aße 18, 40474 Düsseldo , Ge many.
Email: Jan-Beh inge @boeckle .de. Till an T eeck, Ins i u e o Socio-Economics, Uni e si y o Duisbu g-Essen,
Lo ha s aße 65, 47057 Duisbu g, Ge many. Email: ill. [email p o ec ed]. Vincen Vic o , PhD S uden a he
Ins i u e o Socio-Economics, Uni e si y o Duisbu g-Essen, Lo ha s aße 65, 47057 Duisbu g, Ge many. Email:
[email p o ec ed]. We would like o hank he Mac oeconomic Policy Ins i u e (IMK) o inancial sup-
po .
1
1 In oduc ion
We analyze he ole o amily i ms in he all o he labo sha e and he ise o co po a e sa ing
in Ge many. Fo his pu pose, we use a newly de eloped da abase wi h de ailed in o ma ion abou
he ole o con olling amilies in he owne ship and co po a e go e nance s uc u es in publicly
lis ed i ms in Ge many h oughou he pe iod 1993-2019. Ou Family Owne ship and Go e -
nance (FOG) da abase p o ides a comp ehensi e basis o analyzing he in luence o amilies in
lis ed i ms in e ms o con ol igh s and in he wo- ie boa d sys em (managemen boa d and
supe iso y boa d) ha cha ac e izes he Ge man co po a e sec o . While we hope ha he FOG
da abase will be use ul o a a ie y o esea ch ques ions, in his pape we combine i wi h inan-
cial s a emen da a o s udy ends in he labo sha e and co po a e sa ing in Ge many. This ocus
is mo i a ed by he obse a ion ha o e he pe iod 1993-2019 he labo sha e ell and sa ing ose
in b oad segmen s o he publicly lis ed co po a e sec o in Ge many (see Figu e 1 a-b), and ha
hese ends we e much mo e p onounced among amily i ms han non- amily i ms (see Figu e
2 a-b).
Recen li e a u e has documen ed ha co po a e labo sha es ha e allen while co po a e sa -
ing, i.e., co po a e p o i s e ained wi hin i ms, has inc eased as a sha e o co po a e alue added
since he 1980s (Ka aba bounis and Neiman, 2014; Chen e al., 2017). The exis ing li e a u e so
a has ocused mos ly on he global le el and he U.S., bu i has paid li le a en ion o coun y-
speci ic he e ogenei y. Acco ding o Chen e al. (2017), bo h he all in labo sha es and he ise
in co po a e sa ing we e pe asi e ac oss i ms o di e en age and size and loca ed in di e en
coun ies and unlikely o be d i en by s uc u al change o idiosync a ic changes in he ma ke
powe o pa icula i ms o indus ies. Au o e al. (2020), by con as , a ibu e he all o he
labo sha e in he U.S. o he eme gence o a small numbe o “supe s a i ms”. These i ms a e
cha ac e ized by high p oduc i i y and low labo sha es and managed, in a con ex o “winne
akes mos ” ma ke s, o expand hei ma ke sha es conside ably, esul ing in a decline o he ag-
g ega e labo sha e. Keh ig and Vincen (2021) conclude ha he dec ease in he labo sha e in he
U.S. manu ac u ing sec o is be e explained by he ecu ing phenomenon o “ ising s a s”, i.e.,
es ablishmen s whose labo sha e ell ( empo a ily) as hey g ew in size.
Accoun ing o coun y-speci ic cha ac e is ics is, howe e , o g ea impo ance since co po-
a e sec o s ope a e unde di e en ins i u ional se ings in di e en coun ies. In pa icula , he
p e alence o amily i ms whe e a con olling amily domina es he owne ship s uc u e, as op-
posed o i ms wi h a dispe sed owne ship s uc u e, is known o be highe in some coun ies han
in o he s (La Po a e al., 1999; Faccio and Lang, 2002; Aminada and Papaioannou, 2020). Mo e-
o e , he e is some indica ion in p e ious li e a u e ha amily i ms pay lowe wages (S ae and
2
Thesma , 2007) and hold mo e cash (Ande son e al., 2012) han non- amily i ms. In con as o
he U.S. co po a e sec o , ea u ing apid s uc u al change d i en by ela i ely young and highly
dynamic “supe s a s” o “ ising s a s”, Ge many’s co po a e sec o has long been cha ac e ized by
highly specialized, es ablished i ms especially in he expo -o ien ed manu ac u ing sec o . These
i ms ypically pu sue a mo e inc emen al inno a ion model and a e o en con olled by indi idual
amilies, no uncommonly o e se e al gene a ions. While he deg ee o amily con ol a ies in
e ms o sha e owne ship and in ol emen in co po a e go e nance, he b oad uni e se o amily
i ms includes no only ela i ely small and highly specialized niche p oduce s, o en e e ed o
as “hidden champions” (Simon, 2022), bu also global playe s such as BMW (con olled by he
Quand amily since he 1960s) o Me ck (con olled by he Me ck amily since he 1820s).
In his pape , we combine inancial s a emen da a ob ained om he Wo ldscope da abase wi h
in o ma ion on co po a e owne ship and go e nance including 929 non- inancial, publicly lis ed
i ms in Ge many o e he pe iod 1993-2019. To cons uc ou own FOG da abase, we supplemen
in o ma ion ex ac ed om Hoppens ed Ak ien üh e , an es ablished and use ul, albei a he in-
comple e, e e ence wo k wi h in o ma ion on Ge man lis ed i ms, wi h a wide ange o manually
collec ed in o ma ion abou sha eholde s, documen ed in he Online Appendix. We de ine amily
i ms as hose i ms whe e a con olling amily holds a leas 25 pe cen o he o ing igh s, includ-
ing h ough complex py amid s uc u es. D awing on a a ie y o online sou ces, we ca ego ize
sha eholde s based on hei cha ac e is ics, including hei amily connec ions, hei ela ionship
o he company ounde and hei le el o in ol emen in he managemen and supe iso y boa ds.
This app oach allows us no only o iden i y amily-con olled i ms bu o delinea e di e en sub-
g oups o amily i ms based on hei managemen s uc u e, including ounde -managed i ms,
i ms managed by ano he amily membe and ex e nally managed i ms. We can hus documen
owne ship and go e nance s uc u es in he Ge man co po a e sec o o e a pe iod o almos h ee
decades, p o iding a le el o g anula i y no a ailable in he exis ing li e a u e, which o en ocuses
on speci ic poin s in ime (e.g. Aminada and Papaioannou, 2020).
Ou main indings a e as ollows. We ejec bo h he “supe s a scena io” and he “ ising s a
scena io” as explana ions o he e olu ion o he labo sha e among public i ms in Ge many.
Whe eas p e ious wo k o he U.S. sugges s ha he all in he agg ega e labo sha e was due
almos exclusi ely o a small ac ion o as -g owing and e en ually e y la ge i ms o es ablish-
men s wi h e y low labo sha es, we ind ha he all o he labo sha e was a mo e widesp ead
phenomenon among public i ms in Ge many. Based on an analysis o he changing dis ibu ions
o labo sha es and alue added as well as coun e ac ual exe cises, we ind ha he all o he
agg ega e labo sha e in ou sample is no due o an inc ease in ma ke sha e o i ms wi h ini ially
low labo sha e. Ra he , ou da a a e mo e compa ible wi h a “cons an ma ke sha e scena io”,
3

i.e., he all in he agg ega e labo sha e was d i en by he d op in he indi idual labo sha es o
incumben i ms. In pa icula , he labo sha e ell especially s ongly among amily-con olled
i ms. While we do no a emp o iden i y he causal mechanisms behind hese ends, one in e -
es ing inding is ha he labo sha e ell pa icula ly sha ply in sec o s whe e amily businesses a e
mos s ongly ep esen ed, especially in manu ac u ing. Looking a indi idual i m-le el ends,
we also ind ha he ise in co po a e p o i s as a sha e o alue added, implied by a all in he
labo sha e, is mi o ed almos one-by-one in inc eases in co po a e sa ing as a sha e o alue
added, gi en ha axes, in e es paymen s and p o i payou s (di idends plus sha e buybacks) ha e
emained oughly cons an as sha es o alue added. In o he wo ds, he s onge ise o co po a e
sa ing in amily i ms as compa ed o non- amily i ms is due mainly o mo e p onounced alls
in he labo sha e, a he han di e ences in p o i e en ion o payou policies. We also ind ha
highe p o i s did no sys ema ically ansla e in o highe in es men spending so ha co po a e ne
lending (sa ing minus in es men ) inc eased mo e o i ms wi h mo e s ongly dec easing labo
sha es and mo e o amily i ms han non- amily i ms. These esul s a e con i med by eg ession
analyses in which we con ol o a ange o o he i m cha ac e is ics in addi ion o ou amily
owne ship and go e nance a iables. We ind ha amily i ms, as de ined by con en ional own-
e ship h esholds, on a e age ha e lowe labo sha es, highe sa ing and highe ne lending han
non- amily i ms. We can also disen angle he e ec s o di e en le els o amily in ol emen in
co po a e go e nance. In e es ingly, labo sha es a e lowe and co po a e sa ing is highe o am-
ily i ms whe e he con olling amily is ep esen ed ei he on he managemen boa d only, o on
he supe iso y boa d only, o on bo h boa ds, han o amily i ms wi hou amily in ol emen in
ei he co po a e boa d. Mo eo e , labo sha es a e lowe , and co po a e sa ing is highe in hose
amily i ms whe e he ounde is he CEO han in amily i ms led by ano he amily membe .
Ex e nally managed amily i ms s ill ha e lowe labo sha es and highe sa ing han non- amily
i ms, bu highe labo sha es and lowe sa ing han amily i ms whe e he CEO is he ounde o
ano he amily membe , con olling o o he i m cha ac e is ics.
O e all, ou esul s quali y he conclusions by Ka aba bounis and Neiman (2014) and Chen
e al. (2017), ha he obse ed ends in he labo sha e and co po a e sa ing a e pe asi e ac oss
coun ies, ypes o economic ac i i ies and sec o s. Ra he , ou indings sugges ha seemingly
simila de elopmen s a ound he wo ld a e d i en by di e en unde lying mechanisms. In Ge -
many, amily owne ship is a key ac o accoun ing o he all in he labo sha e and he ise in
co po a e sa ing, especially in he manu ac u ing sec o .
An ob ious ques ion o ask is, why do amily businesses manage o pay lowe wages han non-
amily i ms? In an en i onmen o labo ma ke de egula ion and ising employmen insecu i y,
he wage discoun ha employees a e willing o accep o employmen secu i y likely inc eased.
4
Consis en wi h ou indings, i has been a gued ha amily i ms en e in o implici con ac s
wi h hei employees p o iding job secu i y, which may lead employees o accep lowe wages
(Muelle and Philippon, 2011; Bassanini e al., 2013). This a gumen is suppo ed by e idence
which sugges s ha employmen in amily i ms is indeed less sensi i e o shocks (Lee, 2006;
S ae and Thesma , 2007; Bjugg en, 2015). Howe e , we do no ind any s aigh o wa d e idence
o mo e p onounced labo hoa ding in amily i ms in he ace o sales shocks in ou sample.
Consequen ly, employees’ expec a ion o highe employmen s abili y in e u n o lowe wages
may be illuso y, al hough mo e esea ch is necessa y o explo e his ques ion.
F om a mac oeconomic pe spec i e, a limi a ion o ou s udy is ha ou da ase comp ises only
publicly lis ed i ms. Howe e , i is o en emphasized ha he backbone o he Ge man co po a e
sec o a e p i a e, amily- un companies ( he so-called Mi els and), especially in he manu ac-
u ing sec o , which ange om e y small o e y la ge companies. I he di e en labo sha e
and sa ing pa e ns documen ed in his pape ex end o he en i e co po a e sec o , his would
s eng hen he mac oeconomic ele ance o ou indings. P e ious li e a u e sugges s ha he all
in he labo sha e and he ise in co po a e sa ing, which esul ed in a pe sis en ly posi i e ne
lending posi ion o he co po a e sec o , ha e been closely ela ed o he eme gence o Ge many’s
cu en accoun su plus since he ea ly 2000s (IMF, 2017; Beh inge and an T eeck, 2018, 2022).
F om a supply-side pe spec i e, he p ominen posi ion o adi ional amily i ms wi h inc emen-
al inno a ion models in he expo -o ien ed manu ac u ing sec o aises he ques ion o he ex en
o which he Ge man co po a e sec o is capable o apid s uc u al change and a eo ien a ion
owa ds he domes ic economy, which could be wa an ed in iew o cu en ecological, echno-
logical and geos a egic challenges.
The emainde o he pape is s uc u ed as ollows. Sec ion 2 p o ides a e iew o he ela ed
li e a u e. Sec ion 3 p esen s ou da ase .1Sec ion 4 examines he all in he labo sha e, and
Sec ion 5 ocuses on he ise in co po a e sa ing. Sec ion 6 o e s some addi ional discussion
and concludes. An ex ensi e documen a ion o he p ocedu es and sou ces used o cons uc he
da ase is p o ided in an Online Appendix.2
2 Rela ed li e a u e
Ou pape is ela ed o h ee b oad opics in he li e a u e: amily owne ship and go e nance, he
labo sha e, and co po a e sa ing.
Fo a long ime, a common pe cep ion was ha (la ge) publicly lis ed i ms ypically exhibi
1The FOG da abase is a ailable a h ps://www.boeckle .de/pd /imk_wp_ og_da abase.xlsx.
2The Online Appendix is a ailable a h ps://www.boeckle .de/pd /imk_wp_ amily_ i ms_online_
appendix.pd .
5
dispe sed owne ship s uc u es (Be le and Means, 1932). Ye , beginning wi h La Po a e al.
(1999), nume ous s udies ha e shown ha a sizable ac ion e en o la ge public i ms ac ually
a e con olled by indi idual sha eholde s, including amilies. This inding has been con i med
o di e en coun ies, including Ge many, and ime pe iods (e.g. Faccio and Lang, 2002; Lins
e al., 2013; Aminada and Papaioannou, 2020). While he exis ing li e a u e has p ima ily o e ed
de ailed documen a ion o owne ship and go e nance s uc u es in he Ge man co po a e sec o o
speci ic poin s in ime, he owne ship and go e nance da abase used in he p esen pape ex ends
he documen a ion o a pe iod o almos h ee decades.
A la ge body o li e a u e has analyzed a ious aspec s o i m pe o mance in amily i ms.
Al hough indings a y, he e is some consensus ha amily i ms do no pe o m wo se han non-
amily i ms in e ms o p o i abili y (Demse z and Villalonga, 2001; Ande son and Reeb, 2003;
S ae and Thesma , 2007) and i m alue (Claessens e al., 2002; Villalonga and Ami , 2006).
The li e a u e also sugges s ha di e en owne ship and con ol s uc u es, and amily owne ship
in pa icula , a e highly ele an o a a ie y o s a egic decisions such as in es men (Ch isman
and Pa el, 2012; Ande son e al., 2012), acquisi ion and di e si ica ion s a egies (Mille e al.,
2009; Gomez-Mejia e al., 2010), downsizing (Block, 2010), employee and execu i e compensa-
ion (Gomez-Mejia e al., 2003; Muelle and Philippon, 2011; Bassanini e al., 2013), employmen
e en ion ollowing shocks (Lee, 2006; S ae and Thesma , 2007), di idend dis ibu ion policy
(Villalonga and Ami , 2006), and cash holdings (Ande son and Hamadi, 2016). Ou pape is also
ela ed o he li e a u e on amily succession, which looks a di e ences in co po a e beha io
and pe o mance be ween ounde -managed, hei -managed, and ex e nally managed amily i ms
(Villalonga and Ami , 2006; Bennedsen e al., 2007; S ae and Thesma , 2007).
In ecen yea s, a g owing li e a u e has discussed he e olu ion o co po a e labo sha es, al-
hough wi hou conside ing he po en ial ole o amily con ol. Elsby e al. (2013) highligh he
ole o in e na ional ade and especially he eloca ion o manu ac u ing om ad anced economies
o China. Ka aba bounis and Neiman (2014) a gue ha he decline o he ela i e cos o capi al
which induced i ms o shi o mo e capi al-in ensi e p oduc ion esul ed in he all o agg ega e
labo sha es. Simila ly, Acemoglu and Res epo (2018) emphasize he impac o echnological
change, which di e en ially a ec s labo and capi al inpu s. This iew is suppo ed by empi i-
cal s udies ha link de elopmen s in he labo sha e o echnology, pa icula ly au oma ion (e.g.
Dao e al., 2017). De Loecke e al. (2020) and Co a ubias e al. (2020) a gue o diminishing
compe i ion as explana ion o he ise in p o i ma k-ups and a co esponding decline in he labo
sha e. While Ka aba bounis and Neiman (2014) ind ha he global decline o co po a e labo
sha es is d i en by changes wi hin, a he han be ween indus ies, Au o e al. (2020) emphasize
he be ween- i m componen in he e olu ion wi hin indus ies. They a gue ha he decline in
6
agg ega e labo sha es can be explained by he ise o “supe s a s”, whe e i ms wi h ini ially low
labo sha es s ongly g ow in size. Keh ig and Vincen (2021) es he supe s a hypo hesis agains
wo al e na i e scena ios: The “big playe scena io”, whe e ini ially la ge i ms educe hei labo
sha es o e ime, and he “ ising s a scena io”, whe e he labo sha e o ini ially small i ms di-
minishes as hey g ow in size. Using es ablishmen -le el da a o he U.S. manu ac u ing sec o ,
hey conclude ha he ising s a scena io bes explains he da a. Bo h he supe s a and he ising
s a explana ions ha e in common ha hey a ibu e he ajec o y o agg ega e labo sha es o
only a small g oup o i ms. The exis ing li e a u e ha dly examines coun y-speci ic ins i u ional
ac o s. Empi ical esea ch using i m-le el da a, on he o he hand, ocuses p edominan ly on he
U.S. We ex end he li e a u e in bo h espec s by examining he co po a e sec o in Ge many and
emphasizing he ole o amily i ms.
The end owa ds highe co po a e sa ing a he global le el has been documen ed, and i s
mac oeconomic implica ions ha e been discussed, based on na ional accoun s da a (e.g. IMF,
2006; And é e al., 2007; IMF, 2017; Dao and Maggi, 2018; Beh inge and an T eeck, 2022).
Complemen ing na ional accoun s da a wi h a la ge in e na ional i m-le el da ase , Chen e al.
(2017) ind ha , wi h axes and in e es paymen s on deb emaining ela i ely s able o e ime as
sha es o alue added, he ise in co po a e sa ing essen ially mi o s he decline in he labo sha e.
Fu he mo e, hey ind ha he inc ease in sa ing shows no pa icula concen a ion wi hin speci ic
ypes o coun ies, indus ies, o i ms. These conclusions a e suppo ed by a gene al equilib ium
model, whe e changes in he cos o capi al, including declines in he eal in e es a e, he p ice o
in es men goods, co po a e income axes, and inc eases in ma kups, may explain he global ise
in co po a e sa ing, ollowing a dec ease in he labo sha e and “s icky” di idend paymen s.
Al hough i seems plausible ha highe p o i e en ion, as opposed o di idend payou s, is mo-
i a ed by addi ional in es men needs, he weak de elopmen o in es men spending despi e high
co po a e sa ings (G ube and Kamin, 2016) aises skep icism abou such explana ions. Howe e ,
he shi in in es men owa ds in angible asse s ha canno be used as colla e al could p omp
companies o make g ea e use o in e nal sou ces o inancing (Fala o e al., 2022). Simila ly, Fan
and Kalemli-Özcan (2015), who analyze Asian economies, associa e sa ing wi h c edi cons ain s
a he i m le el. O he s udies emphasize a p ecau iona y mo i e and a gue ha i ms accumula e
sa ings as a hedge agains isks in imes o heigh ened unce ain y (Ba es e al., 2009; A men e
and Hna ko ska, 2017). In e es ingly, and ele an o he opic o he p esen pape , Ande son and
Hamadi (2016), using da a on Belgian lis ed i ms be ween 1991 and 2006, ind a s ong posi i e
associa ion be ween owne ship concen a ion and cash holding. This may indica e a p ecau iona y
mo i e on he pa o he con olling sha eholde s who highly alue con ol.
7
p opo ion o amily businesses in mos indus ies is a ound 40 pe cen , al hough he e a e some
ou lie s anging om 13 and 14 pe cen in, espec i ely, he u ili ies and anspo indus ies, which
gene ally consis o a ew la ge en e p ises, p ima ily non- amily-con olled, o 51 and 57 pe cen
in, espec i ely, cons uc ion and anspo a ion equipmen . Family owne ship is pe sis en ly high
in he manu ac u ing sec o , he adi ional backbone o he Ge man economy.
In e es ingly, while 42 pe cen o all i ms quali y as amily i ms ac oss all indus ies, 20
pe cen o all i ms a e amily i ms wi h amily ep esen a ion on he managemen boa d, and 16
pe cen a e amily i ms wi h amily ep esen a ion on he supe iso y boa d, bu in only 4 pe cen
o all i ms he e is a con olling amily which is also ep esen ed on bo h he managemen boa d
and he supe iso y boa d. 9 pe cen o all i ms a e ounde -led amily i ms, 10 pe cen a e
amily i ms whe e ano he amily membe is CEO, and 23 pe cen o all i ms a e amily i ms
wi h an ex e nal CEO.
Table 2 shows key inancial indica o s o he i ms in ou sample, using a e ages o e he
pe iod 1993-2019. The i s se en ows show he unweigh ed mean o he di e en componen s
o g oss alue added o all i ms in ou sample, non- amily and amily i ms, as well as he h ee
subg oups o amily i ms de ined by hei CEO. The emaining ows show unweigh ed a e ages
o o al asse s, i m age, le e age a io, R&D in ensi y, and s ock p ice ola ili y. Wi h asse s o
almos ou billion eu os and an a e age age o sligh ly mo e han 70 yea s, he companies a e,
on a e age, qui e la ge and long-es ablished. Rega ding di e ences in o al asse s be ween i m
ypes, amily i ms a e, on a e age, somewha smalle , due o he smalle size o i ms led by hei
ounde o o he amily membe s. Founde -led companies a e also no ably younge , which is no
su p ising gi en he gene ally ad anced age o he companies in he sample. Howe e , ex e nally
led companies exhibi an a e age size close o non- amily i ms and a e, on a e age, e en a bi
olde han non- amily i ms. This sugges s ha while many la ge i ms a e also amily-owned,
ope a ional managemen is ypically ou sou ced o p o essional manage s abo e a ce ain size.
Ac oss all i ms ypes, a ound h ee-qua e s o g oss alue added is alloca ed o wages and
sala ies on a e age o e he pe iod 1993-2019. Smalle po ions a e used o in e es paymen s,
axes, and dis ibu ions o sha eholde s. 15 pe cen emains wi hin i ms as g oss sa ing. On
a e age ac oss all companies and yea s, in es men exceeds g oss sa ing, esul ing in companies
being, on a e age, ne bo owe s. Family i ms exhibi highe sa ings han non- amily i ms:
While hei labo sha e is a bi mo e han 5 pe cen age poin s lowe , amily i ms pay only sligh ly
mo e in in e es s, axes and payou s as a pe cen age o g oss alue added, so ha sa ing as a
pe cen age o g oss alue added is a bi mo e han 3 pe cen age poin s highe han o non- amily
i ms.
Table 3 shows 10-yea ends o labo compensa ion and co po a e sa ing as a pe cen o
14

g oss alue added o he whole c oss-sec ion o i ms and a he indus y-le el. As shown in
Figu e 1, he all in he labo sha e and concomi an ise in co po a e sa ing is exclusi e o amily
i ms. Non- amily i ms show no only a less p onounced labo sha e end bu also a nega i e
sa ing sha e end. No ably, when examining speci ic indus ies, i s ands ou ha among non-
amily i ms only manu ac u ing subsec o s exhibi s a is ically signi ican ends aligned wi h he
agg ega e de elopmen s o bo h measu es. In con as , all o he indus ies demons a e opposing
ends o lack signi ican ends al oge he . Con e sely, amily i ms show signi ican ends in he
expec ed di ec ions no only ac oss manu ac u ing subsec o s bu also wi hin o he indus ies.
In he nex wo Sec ions, we use ou da ase o in es iga e in g ea e dep h he ole o amily
owne ship and go e nance in he e olu ion o he co po a e labo sha e and co po a e sa ing.
4 Why did he labo sha e all?
4.1 The e olu ion o he labo sha e dis ibu ion
Figu e 1 a) shows di e ging ends in he labo sha es a he agg ega e and i m le el: while
he agg ega e labo sha e o all publicly lis ed non- inancial i ms on a e age declined by 3.7
pe cen age poin s pe decade (see Table 3), he median labo sha e ell a less. This implies
ha he all in he labo sha e is no a uni o m end a ec ing he en i e popula ion o public
non- inancial i ms, bu ha he e was a ealloca ion o alue added owa d he le ail o he
labo sha e dis ibu ion. As discussed by Keh ig and Vincen (2021) in he con ex o he U.S.
manu ac u ing labo sha e, h ee scena ios po en ially could accoun o such a pa e n: in he
“big playe scena io”, ini ially la ge i m educe hei labo sha e o e ime; in he “supe s a
scena io”, i ms wi h an ini ially low labo sha e g ow; and in he “ ising s a scena io”, ini ially
small i ms dec ease hei labo sha e as hey g ow. In he emainde o his Subsec ion, we assess
he ele ance o he h ee scena ios highligh ed by Keh ig and Vincen (2021) while also looking
sepa a ely a amily and non- amily i ms.
An impo an di e ence be ween ou i m-le el da a o Ge many and he U.S. manu ac u ing
es ablishmen -le el da a used by Keh ig and Vincen (2021), which is al eady appa en om Fig-
u e 1 a), is ha he all in he labo sha e is a much mo e widesp ead phenomenon in ou sample
compa ed o he U.S. da a. Keh ig and Vincen (2021) show ha he median labo sha e o U.S.
manu ac u ing es ablishmen s inc eased and was essen ially la a he 25 h pe cen ile o he labo
sha e dis ibu ion o e he pe iod 1967-2012. Simila ly, Au o e al. (2020) show ha he median
ma kup o U.S. manu ac u ing i ms was la du ing 1982-2012, while he agg ega e ma kup ell
s ongly, and ha he all in he labo sha e was d i en by he inc easing concen a ion o alue
15
added among a hand ul o i ms in indi idual indus ies. In ou da a, he all in he labo sha e is
e y simila a he 25 h pe cen ile o he labo sha e dis ibu ion and a he agg ega e le el, and
e en he median shows a clea , albei less p onounced, downwa d end. This shows ha he all
in he labo sha e among public i ms in Ge many was no unique o only a ew supe s a s.
In Figu e 3, we di ide he dis ibu ion o labo sha es in o 10 pe cen age poin -wide bins, om
0 pe cen o 120 pe cen o he i s and he inal yea o ou sample pe iod. Fo each labo sha e
bin, we compu e i s sha e o o al alue added and numbe o es ablishmen s o all i ms (panels
a and b), amily i ms (panels c and d) and non- amily i ms (panels e and ). The black lines
in he panels o Figu e 3 display he dis ibu ion o i ms in 1993 (le panels) and 2019 ( igh
panels) along he labo sha e dimension. The ba s in each panel show he dis ibu ion o alue
added agains he labo sha e. As is appa en om panels a and b o Figu e 3, he dis ibu ion o
i ms has become mo e le -skewed o e ime, wi h a la ge mass in 2019 o low-labo -sha e i ms
ela i e o 1993. A he same ime, we obse e s ong ealloca ion o alue added owa d he le
ail o he labo sha e dis ibu ion. Whe eas less han 10 pe cen o alue added was p oduced in
i ms whose labo sha e ell sho o 60 pe cen in 1993, i ms wi h labo sha es below 50 pe cen
accoun ed o oughly hal o o al alue added in 2019. Despi e he appa en ealloca ion o alue
added owa d he low labo sha e spec um, panel b o Figu e 3 shows ha oughly one hi d o all
i ms in 2019 ha e labo sha es in p oximi y (be ween 40 and 60 pe cen ) o he agg ega e labo
sha e. This is in s a k con as o he supe s a phenomenon obse ed in U.S. da a. The di e ences
in he e olu ion o he labo sha e dis ibu ion be ween amily i ms and non- amily i ms a e
s iking. In 1993, he dis ibu ion o i ms and o alue added was e y simila o bo h ypes o
i ms: mos o alue added was p oduced by i ms in he middle o he labo sha e dis ibu ion
(be ween 60 pe cen and 80 pe cen ). Bu in 2019, mo e han h ee qua e s o o al alue added
by amily i ms was gene a ed by i ms wi h a labo sha e below 40 pe cen ; o non- amily i ms,
only abou 40 pe cen o o al alue added was accoun ed o by i ms whose labo sha e is lowe
han 40 pe cen . A he same ime, he dis ibu ion o i ms was somewha mo e le -skewed o
non- amily i ms in 2019.
Nex , in Figu e 4, we compu e coun e ac ual labo sha es o examine he ela i e impo ance
o changes in ma ke sha es ( ealloca ion o alue added) and changes in i m-le el labo sha es.
In he “cons an labo sha e scena io”, we compa e he ac ual labo sha e o a coun e ac ual in
which we keep he i m’s labo sha e equal o i s ini ial alue while allowing i s ma ke sha e o
e ol e o e ime, as is he case in he da a. Fo his exe cise, we ocus on a s ongly balanced panel
be ween 1993 and 2019 since he ini ial ma ke sha e and labo sha e changes o i ms en e ing
o exi ing a e no well de ined. Despi e i s mo e limi ed co e age, he agg ega e labo sha e end
in his s ongly balanced sample looks e y simila o he one in he ull sample, as is appa en
16
om Figu e 4. This sugges s ha mos o he ealloca ion we documen ed ea lie is occu ing
among long-li ed incumben i ms, a he han being d i en by en y and exi . The mo e closely
he coun e ac ual labo sha e, which is de e mined assuming cons an cons an labo sha es a he
i m le el, acks he ac ual labo sha e, he mo e his could be aken as e idence ha he agg ega e
labo sha e is d i en by supe s a i ms which g ew s ongly in size o e he sample pe iod. We
compu e he coun e ac ual agg ega e labo sha e based on he assump ion on cons an i m-le el
labo sha es, λCLS
, as ollows:
λCLS
=X
i
λi,1993ωi (7)
whe e λi,1993 deno es he labo sha e o i m iin 1993 and ωi, he sha e o i m iin agg ega e g oss
alue added in yea . As can be seen in Figu e 4 a), he cons an labo sha e scena io does no
gene a e any no iceable d op in he labo sha e o e he sample pe iod, compa ed o a dec ease
o mo e han 10 pe cen age poin s in he ac ual labo sha e in he balanced sample. This esul
indica es ha he e we e no majo shi s in he ma ke sha es o incumben i ms wi h low labo
sha es which could explain he agg ega e labo sha e de elopmen . We can hus i mly ejec he
supe s a hypo hesis o ou sample.
In he “cons an ma ke sha e scena io”, shown in Figu e 4 b), we keep each i m’s ma ke
sha e cons an while allowing i s labo sha e o change as in he da a. The coun e ac ual agg ega e
labo sha e based on he assump ion o cons an i m-le el ma ke sha es, λCMS
, is compu ed as
ollows:
λCMS
=X
i
λi ωi,1993 (8)
Figu e 4 b) shows ha he cons an ma ke sha e scena io p oduces a coun e ac ual agg e-
ga e labo sha e se ies ha acks he agg ega e labo sha e obse ed in he ac ual da a ex emely
closely. The como emen o he ac ual and he coun e ac ual labo sha es indica es ha e en i
ma ke sha es had no changed a all, he e olu ion o he agg ega e labo sha e would ha e been
e y simila , d i en by changes in indi idual labo sha es o incumben i ms. Keh ig and Vin-
cen (2021) e e o his la e coun e ac ual exe cise as he “big playe scena io”. I is clea ha
la ge i ms d i e he e olu ion o he agg ega e labo sha e unde his scena io, simply because o
hei la ge sha e in o al alue added.8Howe e , as shown in Figu es 1 a) and 3, he all in i m-
le el labo sha es in ou sample is no unique o jus a ew big playe s, bu i is a mo e pe asi e
8O he en la ges i ms in e ms o g oss alue added in 2019, eigh a e al eady in ou sample in 1993. These eigh
i ms alone (Volkswagen, Me cedes, Siemens, BMW, Baye , SAP, F esenius, and BASF) accoun ed o 40 pe cen o
g oss alue added in ou sample in 1993 and 50 pe cen in 2019.
17
phenomenon among Ge man public non- inancial i ms.
4.2 Wages and p oduc i i y
A de ining ea u e o he supe s a phenomenon in he U.S. is ha supe s a i ms ha e low labo
sha es because hey a e highly p oduc i e ( a he han pay below-a e age wages a a e age p o-
duc i i y) (Au o e al., 2020; Keh ig and Vincen , 2021). Because ou discussion so a sugges s
ha a key ac o in accoun ing o he all in he labo sha e in Ge many a e di e ences be ween
amily i ms and non- amily i ms, we ake a close look a he ela i e con ibu ions o wages
and p oduc i i y o he obse ed labo sha e ends. Ideally, we would compa e hou ly wages and
hou ly labo p oduc i i y ac oss di e en i m ypes. Howe e , Wo ldscope only p o ides i m-
le el in o ma ion abou employmen , no abou hou s wo ked. This is an impo an ca ea o he
discussion ha ollows.
Figu e 5 shows ha amily i ms in he agg ega e had simila labo p oduc i i y as non- amily
i ms, bu paid lowe wages h oughou he en i e sample pe iod. In Table 4, we show esul s o
i m-le el eg essions o wages pe employee and g oss alue added pe employee (labo p oduc-
i i y) on a amily i m dummy as well as indus y and ime ixed e ec s (columns 1 and 3). In
columns 2 and 4 o Table 4, we also con ol o a measu e o capi al in ensi y (asse s o e em-
ploymen ) and a measu e o echnological ad ancemen (R&D expendi u es o e sales), bo h o
which should be posi i ely ela ed wi h p oduc i i y and wages. In e es ingly, amily i ms pay
lowe wages pe employee, bu do no ha e lowe labo p oduc i i y, e en when con olling o
capi al in ensi y and R&D expendi u es. Acco ding o ou eg essions, amily i ms on a e age
pay 8.2 pe cen lowe wages pe employee han non- amily i ms, and 6.6 pe cen lowe wages
pe employee a gi en capi al in ensi y and R&D expendi u es, whe eas he e a e no signi ican
di e ences o p oduc i i y ac oss i m ypes. This esul once again con as s wi h he supe s a
phenomenon obse ed in he U.S.
4.3 The labo sha e and di e en o ms o amily con ol
We also use ou da ase o a se ies o eg ession analyses o examine he implica ions o a ious
deg ees o amily in ol emen in owne ship and go e nance o he labo sha e, while con olling
o o he i m cha ac e is ics. In he nex Sec ion, we use he same eg ession analysis amewo k
o look a he ela ionship be ween amily con ol and he uses o g oss alue added including
co po a e sa ing. The es ima ion equa ions ha e he ollowing gene al o m:
yi =α+βFi +γXi +εi (9)
18
whe e yi is he dependen a iable o i m ia ime (i.e. he labo sha e in his Sec ion, and
di e en componen s o g oss alue added in Sec ion 5).9Fi is ou amily i m a iable, which
can be b oken down in o di e en dummy a iables ep esen ing di e en o ms o amily con ol
depending on he deg ee o amily in ol emen in owne ship and go e nance. Since amily con ol
shows li le a ia ion o e ime, we canno iden i y i m ixed e ec s wi h his es ima ion. Xi is
a se o con ol a iables. These a iables include he log o o al asse s and he log o yea s
since ounda ion o accoun o possible e ec s o i m size and age. Fu he con ol a iables
a e he le e age a io, R&D in ensi y, and s ock p ice ola ili y. These a iables a e s anda d
con ols used in he li e a u e ha analyzes a ious aspec s o i m beha io and pe o mance (e.g.
Ande son and Reeb, 2003; S ae and Thesma , 2007; Lins e al., 2013; Fan and Kalemli-Özcan,
2015; Dao and Maggi, 2018). All es ima ions include indus y and yea ixed e ec s.10 S anda d
e o s a e clus e ed a he i m le el.
The esul s om OLS eg essions a e p esen ed in Tables 5-7. In his Sec ion, we discuss
he esul s o h ee eg essions whe e he labo sha e is he dependen a iable, shown in he i s
column o Tables 5-7. In Table 5, we eg ess i m-le el labo sha es on a amily i m dummy,
whe e amily i m is de ined solely by owne ship ( he amily sha e exceeding 25 pe cen o con-
ol igh s), in addi ion o he con ol a iables.11 We ind ha he labo sha e on a e age is 9.6
pe cen age poin s lowe in amily i ms han in non- amily i ms. In Table 6, we di e en ia e
be ween ou ypes o amily i ms: in i ms whe e he con olling amily is p esen only in he
managemen boa d, he labo sha e is es ima ed o be 12.6 pe cen age poin s lowe han in non-
amily i ms; in i ms whe e he con olling amily is p esen only in he supe iso y boa d, he
labo sha e is 8.3 pe cen age poin s lowe ; in amily i ms wi hou amily ep esen a ion on ei he
boa d, he labo sha e is es ima ed o be s ill 5.6 pe cen age poin s lowe han in non- amily i ms;
when he con olling amily is p esen in he supe iso y boa d in addi ion o being ep esen ed on
he managemen boa d, his does no ha e an addi ional e ec on he labo sha e acco ding o ou
es ima ions. Finally, in Table 7, we look a he implica ions o di e en CEO ypes o amily i ms
o he labo sha e. We ind he s onges e ec o ounde -led amily i ms, whe e he labo sha e
is es ima ed o be 14.7 pe cen age poin s lowe han in non- amily i ms, con olling o o he i m
cha ac e is ics. In amily i ms which a e led by a amily membe o he han he i m’s ounde
he labo sha e is 11.5 pe cen age poin s lowe han in non- amily i ms, while in amily i ms
managed by an ex e nal CEO, he di e ence is 6.9 pe cen age poin s, con olling o o he i m
cha ac e is ics.
9We winso ize all dependen a iables a he op and bo om 1 pe cen .
10We use SIC codes om Wo ldscope o g oup companies in o wel e indus ies, i e o which a e manu ac u ing
subsec o s. Table 3 and Table A.5 in he Online Appendix p o ide an o e iew o he indus y classi ica ion.
11The esul s a e obus o using he owne ship sha e o he con olling amily ins ead o he amily i m dummy.
19

5 How we e ising p o i s used?
In his Sec ion, we analyze how he all o he labo sha e was ansla ed in o a ise o co po a e
sa ing as a sha e o g oss alue added. A he agg ega e le el, we documen ed a end whe e
labo compensa ion has been educed p ima ily in a o o g oss sa ing, especially among amily
i ms (Table 3). This obse a ion p omp s he ques ion o whe he his ela ionship holds a he
indi idual i m le el, i.e., whe he indi idual i ms expe iencing a d op in hei labo sha e co -
espondingly inc eased hei sa ing as a sha e o g oss alue added. I he connec ion be ween
dec easing labo and inc easing sa ing sha e pe sis s a he i m-le el, a la ge (smalle ) end in-
c ease in g oss ope a ing su plus on a e age should be associa ed p ima ily wi h a la ge (smalle )
end inc ease in g oss sa ing a he han payou s, in e es s, o axes.
5.1 Fi ms’ p o i and sa ing ends
As a i s s ep, we plo he pe cen age poin s end pe 10 yea s in i m g oss ope a ing su plus
agains he ends in he ou ca ego ies ha cons i u e i , simila o Chen e al. (2017) in hei
analysis o global co po a e sa ing. The esul s a e shown in Figu e 6, whe e he size o each
ci cle co esponds o he i m’s a e age g oss alue added. Black ci cles ep esen amily i ms,
g ey ci cles non- amily i ms. In p oducing Figu e 6, we es ic he sub-sample o amily i ms o
only hose i ms ha can be iden i ied as amily i ms. S ikingly, only he end in g oss sa ing
exhibi s a s ong posi i e associa ion wi h he end in g oss ope a ing su plus, implying ha i ms
educing hei labo sha e o e he sample pe iod end o di ec addi ional g oss ope a ing su plus
in o g oss sa ing a he han o he po en ial uses. In Figu e 7, we epea he same exe cise o see
how ends in g oss sa ing a e ela ed o in es men and ne lending: i ms can ei he use hei
sa ing o und in es men expendi u es, o o acqui e inancial asse s o educe liabili ies. As is
appa en om Figu e 7, i m-le el ends in g oss sa ing co ela e posi i ely wi h ends in ne
lending, while he e is no disce nible associa ion wi h ends in in es men .
In Figu es 6 and 7, he e is no no iceable di e ence in he end ela ionships be ween amily
i ms and non- amily i ms. We can see some concen a ion o amily i ms in he op- igh seg-
men s o Figu es 6 and 7, implying ha many (incuding la ge) amily i ms expe ienced bo h a all
in he labo sha e and a ise in co po a e sa ing and ne lending. Howe e , he mo e p onounced
ise in co po a e sa ing (and ne lending) among amily i ms seems o be due p ima ily o s onge
d ops in he labo sha e, as opposed o di e ences in he p opensi y o e ain p o i s.
20
5.2 Uses o alue added and di e en o ms o amily con ol
To ake a close look a he di e ences be ween di e en ypes o amily i ms and non- amily
i ms in e ms o uses o alue added, we now u n o he eg ession esul s o Equa ion 9, shown
in Tables 5-7. As no ed p e iously, we con ol o indus y and ime ixed e ec s, i m size (asse s)
and age. We also ake in o accoun he e ec o R&D in ensi y due o i s po en ial in luence on
p oduc i i y and sa ing and in es men decisions. The le e age a io is included o con ol o
deb in he capi al s uc u e. Highe sa ing may be an adap a ion o highe i m speci ic isk, o
which s ock e u n ola ili y is used as a p oxy.
The esul s o he eg ession analyses o ou baseline de ini ion o amily i ms, based on
con ol igh s, a e shown in Table 5: amily i ms use a signi ican ly lowe sha e o g oss alue
added o labo compensa ion which is p ima ily o se by a highe sha e o sa ing. Wi h mo e
han 9 pe cen age poin s lowe labo compensa ion (column 1) and o e 7 pe cen age poin s highe
sa ing (column 5), he magni ude o he di e ences is subs an ial. In e es ingly, we also ind ha
amily i ms ha e signi ican ly highe payou s (column 4). The o e all de elopmen o payou s by
amily i ms shown in p e ious sec ions is p ima ily d i en by la ge amily i ms, which exhibi
excep ionally low payou a ios. We do no ind a signi ican e ec o amily con ol on in es men
(column 6), bu an e ec on ne lending (column 7) o a simila magni ude as he one on g oss
sa ing, indica ing ha amily i ms, despi e ha ing signi ican ly highe in e nal means o inance
a ailable h ough sa ing, do no ha e highe ela i e in es men expendi u es. Ins ead, amily
i ms ha e highe ne lending, all else equal.
In Tables 6 and 7, we show he esul s o eg essions in which di e en subg oups o amily
i ms acco ding o amily in ol emen in managemen a e used as igh -hand side a iables. The
esul s essen ially con i m he main inding om Figu e 6 and Table 5: he mo e o less p onounced
nega i e de ia ions in he labo sha e in he a ious ypes o amily i ms compa ed o non- amily
i ms a e almos en i ely e lec ed in co esponding posi i e de ia ions in co po a e sa ing (and,
hough somewha less clea ly, ne lending) as a pe cen age o alue added.
6 Discussion and conclusion
By emphasizing he ole o amily-con olled i ms as a dis inc i e ea u e o he Ge man co po a e
landscape, his pape complemen s he exis ing li e a u e ha has ended o o e look coun y-
speci ic he e ogenei y in ends o co po a e labo sha es and co po a e sa ing. Ou esul s lea e
a numbe o open ques ions and poin o se e al impo an a enues o u u e esea ch, which we
b ie ly discuss in he emainde o his Sec ion.
21
Fi s , he ques ion emains as o why amily businesses manage o pay lowe wages. Ha e
wo ke s become inc easingly willing o accep lowe wages in exchange o he highe employ-
men secu i y ha amily i ms may be able o c edibly commi o, as sugges ed by he li e a u e
on implici employmen con ac s (Muelle and Philippon, 2011; Bassanini e al., 2013)? Some
p e ious li e a u e sugges s ha employmen in amily i ms is indeed less sensi i e o sales shocks
(Lee, 2006; S ae and Thesma , 2007; Bjugg en, 2015).
While an exhaus i e ea men o his ques ion is beyond he scope o his pape , we do no ind
e idence ha employmen in amily i ms is mo e insula ed om sales shocks. Figu e 8 plo s he
s anda d de ia ions in he log o sales agains he s anda d de ia ions in he log o employmen o
amily and non- amily i ms sepa a ely. Al hough he e is a clea posi i e co ela ion be ween he
s anda d de ia ions, indica ing ha changes in sales a e passed h ough o employees by adjus ing
employmen numbe s, amily i ms and non- amily i ms do no appea o eac di e en ly o
luc ua ions in sales. In Appendix B, we p esen a simple eg ession analysis, whe e employmen
is eg essed on sales, an in e ac ion e m be ween sales and di e en amily i m dummies, and a
numbe o con ol a iables. The esul s sugges ha he esponse o amily i ms o sales shocks
does no di e om ha o non- amily i ms. Hence, i employees we e indeed willing o ade
lowe wages o pe cei ed job secu i y in amily i ms, his ade-o ul ima ely may ha e been
based on a alse assump ion. Ye , Nollenbe ge (2024) p esen s new e idence o a sample o lis ed
and unlis ed i ms ha amily i ms in Ge many educed employmen le els less han non- amily
i ms, con olling o changes in sales and o he i m cha ac e is ics, du ing he ecen Co id and
ene gy p ice c ises. Clea ly, mo e esea ch is needed o in es iga e he ela ionship be ween wages
and employmen s abili y u he .
A second and ela ed ques ion ha equi es mo e esea ch is, why do amily i ms in pa icu-
la e ain such a la ge p opo ion o hei p o i s a he han dis ibu e hem o sha eholde s? We
concluded p e iously ha amily i ms do no di e om non- amily i ms in e ms o payou poli-
cies, bu ha hei highe sa ing esul s p edominan ly om lowe wages ela i e o p oduc i i y.
Con olling amilies may be mo e isk a e se (Ande son e al., 2012) and ha e a highe p e e ence
o in e nal inancing han o he sha eholde s, as hey seek o e ain con ol o hei business and
pass i on o u u e gene a ions. In he Ge man con ex , a p e e ence o accumula e unds inside
i ms may be ampli ied by inhe i ance axa ion which s ongly a o s weal h ied up in companies
(Bach, 2015). I he wo ke s o amily i ms o some ex en sha e hese objec i es and alue he
inancial independence and pe cei ed c isis esilience o hei employe s, his may go some way
in explaining why hey accep lowe wages a gi en p oduc i i y le els.
Thi d, one may ask wha a e he b oade dis ibu ional implica ions o ou indings? While
he ise in e ained p o i s is no accoun ed o in con en ional measu es o pe sonal income dis-
22
ibu ion (which include only household income), in heo y hey should be cap u ed by measu es
o pe sonal weal h dis ibu ion (as sha eholde s a e he ul ima e owne s o co po a ions). Ye , e-
ained p o i s inc ease sha eholde s’ weal h only o he ex en ha hey a e e lec ed in i ms’ s ock
ma ke alue. Hence, s anda d measu es o bo h income and weal h inequali y likely a e down-
wa d biased as a esul o disp opo iona ely high sa ing by i ms wi h concen a ed owne ship
s uc u es.
Fou h, om a mac oeconomic pe spec i e, he impo ance o amily i ms o Ge many’s
b oade “g ow h model” wa an s u he discussion. Fo one hing, he high sa ing a e and pos-
i i e ne lending posi ion o he Ge man co po a e sec o ha e likely con ibu ed o Ge many’s
pe sis en cu en accoun su plus and s uc u al dependence on expo s as a g ow h d i e (IMF,
2017; Beh inge and an T eeck, 2018). A he same ime, he la ge sha e o manu ac u ing in
Ge many’s alue added and he associa ed high deg ee o specializa ion equi e a ocus by co po-
a ions no on he domes ic ma ke , bu on expo s in o de o achie e he necessa y ma ke size.
While he inc emen al inno a ion model o “coo dina ed ma ke economies” such as Ge many has
long been ecognized in he li e a u e on “ a ie ies o capi alism” as an al e na i e o he mo e
adical inno a ion model in “libe al ma ke economies” like he U.S., he ole o amily i ms
in Ge many’s inno a ion and g ow h model so a has been unde app ecia ed in he compa a i e
poli ical economy li e a u e (Beh inge and an T eeck, 2019).
In o de o be e unde s and how he discussion o inno a ion and g ow h models is linked
o he de elopmen o he labo sha e, a dig ession on he “ ising s a ” phenomenon in he U.S.,
highligh ed by Keh ig and Vincen (2021), may be use ul. The ising s a in e p e a ion in ac de-
sc ibes a ecu ing cycle o adical p oduc inno a ion: New i ms de elop new p oduc s ha mee
he as es o consume s. This ansla es in o unusually high demand, high p ice-se ing powe ,
and hence excep ionally high labo p oduc i i y. This, in conjunc ion wi h incomple e wage ad-
jus men s, esul s in a low labo sha e. Tempo a ily high p o i sha es e en ually decline as he
compe i ion adjus s o s uc u al change and he ising s a s’ p ice-se ing powe alls. In addi ion,
he ising s a s mus g adually pay highe wages and expand employmen . This leads o a ise
in he labo sha e o hese o me ising s a s. Then new ising s a s come along, de elop be e
and al oge he di e en p oduc s, and he cycle begins again o hese new ising s a s while he
old ising s a s’ labo sha es ha e al eady isen again and hei ma ke sha es ha e allen due o
he success o he new ising s a s. The phenomenon o ising s a s explains why mos o oday’s
bes -known U.S. companies, such as Apple, Google, Mic oso , and Amazon, we e comple ely
unknown wo o h ee decades ago.
The si ua ion is e y di e en in Ge many, whe e he co po a e landscape is s ill domina ed
by he same incumben s as in he ea ly 1990s, especially in he au omo i e, chemical, and pha -
23
0
20
40
60
80
in %
0 20 40 60 80 100 120
Labo sha e, in %
Sha e o alue added
Sha e o i ms
(a) All i ms, 1993
0
20
40
60
80
in %
0 20 40 60 80 100 120
Labo sha e, in %
Sha e o alue added
Sha e o i ms
(b) All i ms, 2019
0
20
40
60
80
in %
0 20 40 60 80 100 120
Labo sha e, in %
Sha e o alue added
Sha e o i ms
(c) Family i ms, 1993
0
20
40
60
80
in %
0 20 40 60 80 100 120
Labo sha e, in %
Sha e o alue added
Sha e o i ms
(d) Family i ms, 2019
0
20
40
60
80
in %
0 20 40 60 80 100 120
Labo sha e, in %
Sha e o alue added
Sha e o i ms
(e) Non- amily i ms, 1993
0
20
40
60
80
in %
0 20 40 60 80 100 120
Labo sha e, in %
Sha e o alue added
Sha e o i ms
( ) Non- amily i ms, 2019
No es: The igu e shows he dis ibu ions o i ms and alued added in 1993 (le panels) and 2019 ( igh panels) along
he labo sha e dimension o all i ms ( op ow), amily i ms (middle ow) and non- amily i ms (bo om ow). The
black lines e lec he dis ibu ion o i ms, while he dis ibu ion o alued added is ep esen ed by he g ay ba s.
Figu e 3: Dis ibu ions o labo sha es and alue added
30

45
50
55
60
65
70
in %
1990 1995 2000 2005 2010 2015 2020
Ac ual (bal. panel)
Ac ual ( ull panel)
Coun e ac ual (bal. panel)
(a) Cons an labo sha e scena io
45
50
55
60
65
70
in %
1990 1995 2000 2005 2010 2015 2020
Ac ual (bal. panel)
Ac ual ( ull panel)
Coun e ac ual (bal. panel)
(b) Cons an ma ke sha e scena io
No es: The igu e shows he ac ual agg ega e labo sha e agains he coun e ac ual labo sha es in he cons an labo
sha e scena io (le panel) and he cons an ma ke sha e scena io ( igh panel). Black lines e lec ac ual labo sha es
o a s ongly balanced panel (solid) and he ull panel (dashed), and g ey lines e lec coun e ac ual labo sha es.
Figu e 4: The e olu ion o he labo sha e unde di e en scena ios
30
40
50
60
70
80
in 1,000 Eu o
1990 1995 2000 2005 2010 2015 2020
Family i ms
Non− amily i ms
(a) Wages pe wo ke
40
60
80
100
120
140
in 1,000 Eu o
1990 1995 2000 2005 2010 2015 2020
Family i ms
Non− amily i ms
(b) Value added pe wo ke
No es: The igu e shows agg ega e wages pe wo ke (le panel) and alue added pe wo ke ( igh panel) o amily
and non- amily i ms.
Figu e 5: Wages and labo p oduc i i y
31
−20
−10
0
10
20
T end in sa ing o GVA
−20 −10 0 10 20
T end in g oss ope a ing su plus o GVA
Family i m
Non− amily i m
(a) Sa ing
−20
−10
0
10
20
T end in in e es o GVA
−20 −10 0 10 20
T end in g oss ope a ing su plus o GVA
Family i m
Non− amily i m
(b) In e es
−20
−10
0
10
20
T end in axes o GVA
−20 −10 0 10 20
T end in g oss ope a ing su plus o GVA
Family i m
Non− amily i m
(c) Taxes
−20
−10
0
10
20
T end in payou s o GVA
−20 −10 0 10 20
T end in g oss ope a ing su plus o GVA
Family i m
Non− amily i m
(d) Payou s
No es: The igu e shows sca e plo s o i m-le el ends (in pe cen age poin s pe 10 yea s) in g oss ope a ing su plus
o g oss alue added agains ends in he componen s ha cons i u e g oss ope a ing su plus (i.e. sa ing, in e es ,
axes, payou s) o amily and non- amily i ms. T ends a e es ima ed o i ms wi h a leas 10 yea s o da a. The
classi ica ion o amily and non- amily i ms is based on he s a us ha occu ed mos equen ly du ing he
obse a ion pe iod. The size o he hollow ci cles co esponds o a i m’s a e age g oss alued added o e he sample
pe iod.
Figu e 6: Uses o ope a ing su plus: sa ing, in e es , axes, payou s
32
−20
−10
0
10
20
T end in in es men o GVA
−20 −10 0 10 20
T end in sa ing o GVA
Family i m
Non− amily i m
(a) In es men
−20
−10
0
10
20
T end in ne lending o GVA
−20 −10 0 10 20
T end in sa ing o GVA
Family i m
Non− amily i m
(b) Ne lending
No es: The igu e shows sca e plo s o i m-le el ends in sa ing o g oss alue added agains ends in in es men
(le panel) and ne lending ( igh panel) o amily and non- amily i ms. T ends a e es ima ed o i ms wi h a leas
10 yea s o da a. The classi ica ion o amily and non- amily i ms is based on he s a us ha occu ed mos equen ly
du ing he obse a ion pe iod. The size o he hollow ci cles co esponds o a i m’s a e age g oss alued added o e
he sample pe iod.
Figu e 7: Uses o sa ing: in es men and ne lending
0
.5
1
1.5
2
S d. De . Log(Employmen )
0 .5 1 1.5 2
S d. De . Log(Sales)
Family i m
Non− amily i m
No es: The igu e plo s he s anda d de ia ion o he log o sales agains he s anda d de ia ion o he log o
employmen o amily and non- amily i ms. S anda d de ia ions a e calcula ed o i ms wi h a leas 10 yea s o
da a. The classi ica ion o amily and non- amily i ms is based on he s a us ha occu ed mos equen ly du ing he
obse a ion pe iod.
Figu e 8: Sales and employmen ins abili y
33
Table 1: P esence o amily i ms
Family i m, wi h amily on Family i m, managed by
Family i m MB SB MB and SB Founde Membe Ex e nal CEO
Indus y (1) (2) (3) (4) (5) (6) (7)
To al 0.42 0.20 0.16 0.04 0.09 0.10 0.23
Ag icul u e and mining 0.32 0.21 0.19 0.13 0.12 0.08 0.12
Cons uc ion 0.51 0.21 0.26 0.00 0.07 0.14 0.30
In o ma ion and communica ions 0.44 0.28 0.13 0.04 0.23 0.05 0.17
To al manu ac u ing 0.45 0.19 0.21 0.05 0.06 0.12 0.27
Chemical and pha maceu ical 0.48 0.20 0.22 0.04 0.07 0.12 0.29
Elec onics 0.43 0.19 0.17 0.03 0.10 0.10 0.24
T anspo a ion equipmen 0.57 0.11 0.38 0.01 0.06 0.05 0.46
Rubbe , plas ic, glass, me al 0.41 0.15 0.25 0.07 0.02 0.10 0.29
O he manu ac u ing 0.46 0.22 0.18 0.06 0.06 0.15 0.25
Se ices 0.38 0.22 0.08 0.02 0.11 0.11 0.16
T anspo a ion 0.14 0.05 0.01 0.00 0.05 0.00 0.08
U ili ies 0.13 0.06 0.06 0.00 0.05 0.01 0.07
Wholesale/Re ail ade 0.47 0.20 0.10 0.04 0.09 0.10 0.28
No es: The able p esen s he sha e o di e en g oups o amily i ms by indus y. Column 1 shows he sha e o amily i ms, columns 2-4 show
he sha e o amily i ms wi h amily ep esen a ion on he managemen boa d (MB), he supe iso y boa d (SB), o bo h boa ds, and columns 5-7
show he sha e o amily i ms de ined by hei CEO.
Table 2: Cha ac e is ics o i ms
Family i m, managed by
All i ms Non- amily i ms Family i ms Founde Membe Ex e nal CEO
(1) (2) (3) (4) (5) (6)
Sala y in % o GVA 73.31 75.46 70.35 69.73 68.80 71.26
In e es in % o GVA 3.05 2.90 3.27 3.46 3.50 3.09
Taxes in % o GVA 4.09 3.70 4.63 5.38 4.99 4.16
Payou s in % o GVA 4.08 3.91 4.31 5.04 4.67 3.85
Sa ing in % o GVA 14.99 13.66 16.84 15.24 17.76 17.10
In es men in % o GVA 22.19 22.73 21.45 25.37 18.54 21.10
Asse s (bn. eu o) 3.94 4.98 2.50 0.26 0.36 4.33
Age (yea s) 71.09 74.28 66.71 22.28 75.51 81.06
Deb /Asse s 20.21 19.09 21.75 19.51 23.96 21.72
R&D/Sales 2.42 2.39 2.46 3.16 1.33 2.66
S ock p ice ol. 4.96 5.14 4.70 5.48 3.56 4.88
Employmen 11849.64 15164.65 7264.65 1182.66 2272.57 11844.91
Sales (bn. eu o) 3.06 3.87 1.94 0.24 0.42 3.28
No es: The able p esen s summa y s a is ics o he di e en componen s o alue added and o he key a iables o all i ms (column
1), non- amily and amily i ms (columns 2 and 3), as well as he h ee g oups o amily i ms de ined by hei CEO (columns 4-6).
34
Table 3: T ends in labo and sa ing sha es
All i ms Family i ms Non- amily i ms
Sala y Sa ing Sala y Sa ing Sala y Sa ing
Indus y (1) (2) (3) (4) (5) (6)
To al -3.690*** 0.665 -8.252*** 7.613*** -1.819 -2.209**
(0.084) (0.054) (0.086) (0.076) (0.099) (0.075)
Ag icul u e and mining -1.058 -2.624 -9.411*** 11.210*** 2.012 -8.730*
(0.244) (0.193) (0.079) (0.098) (0.338) (0.376)
Cons uc ion -2.781 -0.367 -10.271*** 5.322*** -1.763 -1.379
(0.163) (0.135) (0.163) (0.131) (0.211) (0.193)
In o ma ion and communica ions -3.276 -2.259 0.745 -4.955*** -3.565* -2.083
(0.163) (0.125) (0.117) (0.103) (0.171) (0.132)
To al manu ac u ing -6.261*** 3.922*** -8.632*** 8.126*** -4.888*** 0.872
(0.085) (0.063) (0.133) (0.110) (0.085) (0.070)
Chemical and pha maceu ical -5.004*** 2.146*** -7.028*** 6.809*** -4.711*** 0.976*
(0.080) (0.045) (0.089) (0.071) (0.083) (0.047)
Elec onics -9.773*** 7.336*** -9.344*** 7.963*** -9.773*** 6.788***
(0.123) (0.099) (0.163) (0.191) (0.132) (0.112)
T anspo a ion equipmen -6.143*** 3.700*** -6.269*** 5.764** -5.203*** 0.488
(0.086) (0.096) (0.168) (0.170) (0.113) (0.132)
Rubbe , plas ic, glass, me al -6.583*** 3.895*** -11.621*** 8.959*** -2.461* -1.571
(0.107) (0.073) (0.115) (0.097) (0.105) (0.089)
O he manu ac u ing -3.433*** 1.733* -3.256*** 2.690*** -3.547** 1.336
(0.089) (0.078) (0.057) (0.058) (0.126) (0.110)
Se ices 2.232** -2.567** -7.578*** 4.469*** -1.098 0.586
(0.067) (0.072) (0.122) (0.115) (0.097) (0.109)
T anspo a ion 1.420 -3.056** 1.420 -0.981 1.534 -3.092**
(0.123) (0.103) (0.650) (0.525) (0.125) (0.104)
U ili ies 0.533 -7.720 -11.205* 2.913 0.764 -7.872
(0.425) (0.434) (0.430) (0.450) (0.438) (0.444)
Wholesale/Re ail ade 2.449** -4.455*** 3.025* -3.854*** 0.758 -3.829***
(0.066) (0.070) (0.123) (0.088) (0.105) (0.067)
No es: The able p esen s ends (in pe cen age poin s pe 10 yea s) in labo and sa ing sha es, a he agg ega e and indus y le el. Columns 1
and 2 show ends o all i ms, columns 3 and 4 show ends o amily i ms and columns 5 and 6 show ends o non- amily i ms. Robus
s anda d e o s a e in pa en heses. *, **, and *** deno es signi icance a 10%, 5%, and 1% le els, espec i ely.
35

Table 4: Wages and p oduc i i y
Log(Sala y/Empl.) Log(GVA/Empl.)
(1) (2) (3) (4)
Family i m -0.082*** -0.066*** -0.007 0.022
(0.025) (0.021) (0.032) (0.021)
Log(Asse s/Empl.) 0.248*** 0.510***
(0.021) (0.017)
R&D/Sales 0.008*** 0.001
(0.001) (0.003)
Indus y ixed e ec s Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes
Obse a ions 10256 10239 10256 10239
Adj. R-squa ed 0.135 0.299 0.174 0.523
No es: The able p esen s he esul s o OLS eg essions. Dependen a iables a e he log
o wages pe wo ke (columns 1 and 2) and he log o alue added pe wo ke (columns 3
and 4). Family i m is a dummy a iable indica ing amily owne ship. O he explana o y
a iables (columns 2 and 4) a e Log(Asse s/Employmen ) (log o o al asse s o e o al
employmen ) and R&D/Sales (R&D in ensi y). All eg essions include indus y ixed
e ec s and yea ixed e ec s. S anda d e o s a e clus e ed a he i m le el. *, **, and
*** deno es signi icance a 10%, 5%, and 1% le els, espec i ely.
Table 5: Componen s o alue added and he ole o amily i ms
Sala y In e es Taxes Payou s Sa ing In es men Ne lending
(1) (2) (3) (4) (5) (6) (7)
Family i m -9.613*** -0.287 1.420*** 0.989*** 7.264*** -0.414 8.379***
(1.583) (0.228) (0.255) (0.326) (1.540) (1.122) (2.268)
Log(Asse s) -5.199*** -0.179*** 0.433*** 0.420*** 4.548*** 0.024 4.987***
(0.580) (0.067) (0.072) (0.090) (0.572) (0.308) (0.765)
Log(Age) 0.259 -0.356** -0.449*** -0.285* 0.787 -3.516*** 5.545***
(0.955) (0.147) (0.157) (0.172) (0.917) (0.795) (1.441)
Deb /Asse s 0.134** 0.227*** -0.063*** -0.056*** -0.239*** 0.202*** -0.471***
(0.054) (0.010) (0.007) (0.009) (0.054) (0.035) (0.078)
R&D/Sales 0.571** -0.024 -0.121*** -0.052* -0.305 2.371*** -2.837***
(0.286) (0.022) (0.023) (0.030) (0.311) (0.229) (0.532)
S ock p ice ol. 0.079 -0.005 0.028** 0.052*** -0.172** 0.227*** -0.432***
(0.087) (0.008) (0.013) (0.015) (0.084) (0.068) (0.124)
Indus y ixed e ec s Yes Yes Yes Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes
Obse a ions 10148 10148 10148 10148 10148 10148 10148
Adj. R-squa ed 0.118 0.378 0.084 0.098 0.078 0.274 0.118
No es: The able p esen s he esul s o OLS eg essions. Dependen a iables a e he di e en componen s o alue added, in pe cen
o alue added. Family i m is a dummy a iable indica ing amily owne ship. O he explana o y a iables a e Log(Asse s) (log o book
alue o o al asse s), Log(Age) (log o i m age measu ed in yea s plus one), Deb /Asse s (le e age a io), R&D/Sales (R&D in ensi y),
S ock p ice ola ili y (s anda d de ia ion o a i m’s s ock p ice). All eg essions include indus y ixed e ec s and yea ixed e ec s.
S anda d e o s a e clus e ed a he i m le el. *, **, and *** deno es signi icance a 10%, 5%, and 1% le els, espec i ely.
36
Table 6: Componen s o alue added and he ole o boa d ep esen a ion o amily i ms
Sala y In e es Taxes Payou s Sa ing In es men Ne lending
(1) (2) (3) (4) (5) (6) (7)
Family in MB -12.578*** -0.408 2.123*** 1.862*** 9.022*** -0.019 10.103***
(2.204) (0.354) (0.407) (0.499) (2.104) (1.818) (3.104)
Family in SB -8.346*** -0.136 0.975*** 0.725 6.499*** -2.984** 10.010***
(2.083) (0.311) (0.350) (0.523) (1.946) (1.295) (2.855)
Family in MB and SB -12.266*** -0.056 1.601*** 0.879 9.048*** 2.362 7.663*
(2.696) (0.616) (0.515) (0.690) (2.545) (2.530) (4.271)
Family nei he in MB no SB -5.631*** -0.376 0.825** 0.039 4.836** 0.952 4.143
(1.952) (0.269) (0.360) (0.356) (1.985) (1.648) (3.143)
Log(Asse s) -5.275*** -0.182*** 0.451*** 0.439*** 4.594*** 0.059 5.006***
(0.581) (0.068) (0.073) (0.090) (0.573) (0.314) (0.766)
Log(Age) -0.008 -0.363** -0.392** -0.210 0.946 -3.520*** 5.739***
(0.965) (0.148) (0.158) (0.174) (0.929) (0.810) (1.453)
Deb /Asse s 0.137** 0.227*** -0.063*** -0.057*** -0.240*** 0.200*** -0.472***
(0.054) (0.010) (0.007) (0.009) (0.054) (0.035) (0.078)
R&D/Sales 0.552* -0.024 -0.118*** -0.048 -0.294 2.373*** -2.826***
(0.283) (0.022) (0.023) (0.030) (0.310) (0.227) (0.527)
S ock p ice ol. 0.082 -0.005 0.028** 0.050*** -0.173** 0.231*** -0.438***
(0.087) (0.008) (0.012) (0.015) (0.085) (0.068) (0.124)
Indus y ixed e ec s Yes Yes Yes Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes
Obse a ions 10148 10148 10148 10148 10148 10148 10148
R-squa ed 0.124 0.380 0.091 0.106 0.083 0.278 0.122
Adj. R-squa ed 0.120 0.378 0.087 0.102 0.078 0.275 0.118
No es: The able p esen s he esul s o OLS eg essions. Dependen a iables a e he di e en componen s o alue added, in pe cen o alue added.
Family in MB, Family in SB, Family in MB and SB, and Family nei he in MB no SB a e dummy a iables indica ing ha he con olling amily
is ep esen ed ei he on he managemen boa d only, on he supe iso y boa d only, on bo h boa ds, o on nei he boa d. The o he explana o y
a iables a e consis en wi h Table 5. All eg essions include indus y ixed e ec s and yea ixed e ec s. S anda d e o s a e clus e ed a he i m
le el. *, **, and *** deno es signi icance a 10%, 5%, and 1% le els, espec i ely.
37
Table 7: Componen s o alue added and he ole o di e en CEO ypes o amily i ms
Sala y In e es Taxes Payou s Sa ing In es men Ne lending
(1) (2) (3) (4) (5) (6) (7)
Founde CEO -14.704*** 0.194 2.399*** 1.559*** 10.158*** 1.731 9.375**
(3.010) (0.559) (0.560) (0.598) (2.821) (2.190) (4.554)
Membe CEO -11.487*** -0.661** 1.925*** 1.762*** 8.536*** -0.757 10.316***
(2.325) (0.332) (0.449) (0.608) (2.219) (2.126) (2.905)
Ex e nal CEO -6.864*** -0.319 0.832*** 0.448 5.616*** -1.104 7.193***
(1.688) (0.234) (0.274) (0.361) (1.643) (1.169) (2.504)
Log(Asse s) -5.288*** -0.179*** 0.452*** 0.438*** 4.602*** 0.046 5.026***
(0.580) (0.067) (0.072) (0.090) (0.573) (0.313) (0.766)
Log(Age) -0.158 -0.314** -0.370** -0.241 1.023 -3.335*** 5.619***
(0.990) (0.149) (0.162) (0.182) (0.949) (0.789) (1.473)
Deb /Asse s 0.136** 0.227*** -0.063*** -0.057*** -0.240*** 0.202*** -0.473***
(0.054) (0.010) (0.007) (0.009) (0.054) (0.034) (0.078)
R&D/Sales 0.556** -0.024 -0.118*** -0.049 -0.296 2.373*** -2.829***
(0.283) (0.022) (0.023) (0.030) (0.310) (0.231) (0.531)
S ock p ice ol. 0.083 -0.006 0.028** 0.052*** -0.174** 0.224*** -0.431***
(0.087) (0.008) (0.012) (0.015) (0.085) (0.068) (0.124)
Indus y ixed e ec s Yes Yes Yes Yes Yes Yes Yes
Yea ixed e ec s Yes Yes Yes Yes Yes Yes Yes
Obse a ions 10148 10148 10148 10148 10148 10148 10148
Adj. R-squa ed 0.121 0.378 0.088 0.101 0.079 0.274 0.118
No es: The able p esen s he esul s o OLS eg essions. Dependen a iables a e he di e en componen s o alue added, in pe cen o
alue added. Founde CEO, Membe CEO, and Ex e nal CEO a e dummy a iables indica ing ha a amily i m is managed ei he by i s
ounde , ano he amily membe , o an ex e nal CEO. The o he explana o y a iables a e consis en wi h Table 5. All eg essions include
indus y ixed e ec s and yea ixed e ec s. S anda d e o s a e clus e ed a he i m le el. *, **, and *** deno es signi icance a 10%, 5%,
and 1% le els, espec i ely.
38
A Illus a ion o he FOG da abase
The iden i ica ion o owne ship and go e nance s uc u es is illus a ed below using he examples
o wo i ms wi h di e en deg ees o amily in ol emen : Con inen al, a la ge au omo i e supplie
which employed mo e han 240,000 pe sons in mo e han 50 coun ies wo ldwide and eco ded
sales o mo e han 44 billion Eu os in 2019; and G ei enbe ge , a small domes ic p oduce o
me al bandsaw blades and p ecision s ip s eel loca ed in Ba a ia wi h 310 employees and sales o
49 million Eu os in 2019.
To be able o ead he ables in his Appendix, a numbe o a iable de ini ions a e equi ed.
Hoppens ed p o ides in o ma ion on he i s le el o owne ship, i.e., he di ec sha eholde s o
he company. Fo each sha eholde wo main pieces o in o ma ion a e p o ided: he name o
he sha eholde (sh_name) and he pe cen age o sha es held (sh_sha e). Fo some sha eholde s
addi ional in o ma ion o a ious kinds is a ailable h ough Hoppens ed , o o he s we need o
ely on ex e nal sou ces. We add six addi ional a iables, which we de ine as ollows: sha eholde
ype, sh_ ype (1: indi idual/ amily, 2: amily holding/ ounda ion, 3: ins i u ional in es o , 4: i m
in same/ ela ed sec o , 5: miscellaneous, 6: s a e (domes ic), 7: s a e ( o eign), 8: employees,
9: easu y sha es); amily a ilia ion, sh_ am (1: yes, 0: no); ela ion o he ounde , sh_ ound
(1: ounde , 2: ela i e o ounde , 0: no ela ion); CEO, sh_ceo (1: yes, 0: no), membe o
managemen boa d; sh_mgm (1: yes, 0: no); membe o he supe iso y boa d, sh_boa d (1: yes,
0: no). These a iables a e coded manually, based on di e en sou ces o in o ma ion. See he
Online Appendix o a mo e de ailed desc ip ion o de ini ions and sou ces.
A.1 Con inen al (2016)
Figu e A.1 shows he owne ship and go e nance s uc u e o Con inen al AG in 2016. The com-
pany has been con olled by he Schae le amily since 2009, h ough a chain o holding compa-
nies, ul ima ely owning 46% o sha es in 2016. While Con inen al had many di e en sha eholde s
in pas yea s (see Online Appendix), he only o he la ge sha eholde in ecen yea s was Black-
Rock (3.15% in 2016), wi h he emaining sha es in ee loa . In addi ion o hei owne ship s ake,
wo membe s o he Schae le amily a e ep esen ed in he company’s supe iso y boa d.
The da a used o ack his s uc u e is de ailed in Tables A.1-A.5. Table A.1 shows he i s
sha eholde o Con inen al AG included in Hoppens ed o 2016. The i s column con ains he
di e en a iables used o ca ego ize he sha eholde . The second column shows he alue as-
signed o he a iable, he hi d column explains he ca ego iza ion, and he las column con ains
he espec i e sou ce. The name o he sha eholde , Schae le Ve wal ung Zwei GmbH, and he
co esponding owne ship s ake o 35.99% a e indica ed in Hoppens ed . Addi ional in o ma ion
39
Table A.5: Sha eholde 5, Con inen al AG, 2016
Va iable Coding Explana ion Sou ce
sh_name5 Geo g F.W.
Schae le
Added as sha eholde be-
cause pa o he boa d and
holds sha es indi ec ly
Hoppens ed : Name included
in supe iso y boa d da a
sh_sha e5 0 No sha es held di ec ly; in-
cluded in sha eholde s 1 and
2.
Hoppens ed : No included in
sha eholde da a
sh_ ype5 1 Na u al Pe son
sh_ am5 1 Membe o Schae le amily
which is he la ges (cumula-
i e) sha eholde .
Ex e nal sou ce: “Geo g
F ied ich Wilhelm Schae le
[...] is a Ge man billion-
ai e businessman and owne
o 80% o he holding com-
pany [...] His mo he , Ma ia-
Elisabe h Schae le , owns
he o he 20%. Bo h inhe -
i ed hei o une om Scha-
e le ’s a he , Geo g Scha-
e le , who died in 1996.”
h ps://de.wikipedia.o g/
wiki/Geo g_F._W._ Schae le
sh_ ound5 0 Schae le amily no associ-
a ed wi h ounde
Ex e nal sou ce: „Keimzelle
de heu igen Con inen al
AG wa die Konku smasse
des kleinen gummi e a -
bei enden Un e nehmens
Neue Hanno e sche Gummi-
Wa en ab ik, die 1869 om
Bankie Mo i z Magnus
[...] e wo ben wu de.“
h ps://de.wikipedia.o g/
wiki/Con inen al_AG
sh_ceo5 0 No a membe o manage-
men boa d
Hoppens ed : Managemen
boa d da a
sh_mgm 5 0 No a membe o manage-
men boa d
Hoppens ed : Managemen
boa d da a
sh_boa d5 1 Membe o supe iso y
boa d
Hoppens ed : Supe iso y
boa d da a
46

Table A.6: Family i m de ini ions, Con inen al AG, 2016
De ini ion Con inen al AG, 2016
Family sha e Cumula i e con olling sha e
held by membe s o he amily
46%
Owne ship Family sha eholde s hold mo e
han 25%
Yes
Owne ship & managemen
boa d
Mo e han 25% and a leas one
amily membe in managemen
boa d
No
Owne ship & supe iso y
boa d
Mo e han 25% and a leas one
amily membe in supe iso y
boa d
Yes
Owne ship & managemen
& supe iso y boa d
Mo e han 25% and amily mem-
be s in bo h boa ds
No
Owne ship & ounde CEO Mo e han 25% and CEO is he
ounde
No
Owne ship & o he amily
membe CEO
Mo e han 25% and CEO is an-
o he amily membe
No
Owne ship & ex e nal CEO Mo e han 25% and CEO is no a
amily membe
Yes
47
A.2 G ei enbe ge (2013)
Figu e A.2 shows he owne ship and managemen s uc u e o G ei enbe ge AG in 2013. The
company is con olled by he G ei enbe ge amily h ough hei amily holding. One amily
membe , S e an G ei enbe ge , ac ed as CEO and sole membe o he managemen boa d o he
company in 2013, while ano he , Heinz G ei enbe ge , was membe o he supe iso y boa d (in
2016, Heinz G ei enbe ge e i ed om he supe iso y boa d, and S e an G ei enbe ge joined
he supe iso y boa d while lea ing he managemen boa d and being eplaced by an ex e nal
CEO). Wi h a s ake o 58.87%, he i s sha eholde o G ei enbe ge AG included in Hoppens -
ed o 2013 is G ei enbe ge Holding GmbH (Table A.7), which can be iden i ied as a holding
company, in which Heinz G ei enbe ge holds 41.2% and S e an G ei enbe ge holds 29.4%.
Acco dingly, he sha eholde is ca ego ized as associa ed wi h he con olling amily.
The second sha eholde (Table A.8) is an ins i u ional in es o called Baden-Wü embe gische
Ve so gungsans al ü Ä z e con olled by a Ge man s a e en i y and is no associa ed wi h he
G ei enbe ge amily. In 2013, S e an G ei enbe ge ac ed as sole di ec o and Heinz G ei en-
be ge was a membe o he supe iso y boa d. We hus added hese wo indi iduals as sha e-
holde s (Tables A.9-A.10). We assign 0 sha es since all sha es which a e (indi ec ly) owned by
sha eholde s 3 and 4 a e al eady lis ed unde sha eholde 1. Online esea ch e eals ha Heinz
G ei enbe ge is he company’s ounde , while S e an G ei enbe ge is his son. The amily is
ep esen ed in he managemen boa d h ough S e an G ei enbe ge who is also he CEO, and he
amily is also p esen in he supe iso y boa d h ough Heinz G ei enbe ge .
The in o ma ion desc ibed abo e is summa ized in Table A.11. Since amily membe s hold
he o e whelming majo i y o sha es in he holding company G ei enbe ge Holding GmbH, we
a ibu e he en i e s ake o he holding company in G ei enbe ge AG o he amily. The amily
sha e acco dingly amoun s o 58.87%. Since he sha e exceeds 25%, G ei enbe ge AG is a
amily i m acco ding o ou baseline owne ship de ini ion. Mo e p ecisely, i is a amily i m
whe e he con olling amily is ep esen ed on bo h he managemen boa d and he supe iso y
boa d. Mo eo e , i is a amily i m led by a amily membe as CEO.
48
G ei enbe ge AG
Supe iso y boa d
H. G ei enbe ge ; D. Schenk; R. Fö s e ; H.
Langho s ; M. F eihe on Mal zan; H. Ransbe ge
Managemen boa d
S. G ei enbe ge
Baden-Wü embe gische
Ve so gungsans al ü Ä z e
G ei enbe ge
Holding GmbH F ee loa
6.2% 58.87% 34.93%
LBBW Asse
Managemen
Heinz
G ei enbe ge
S e an
G ei enbe ge
100% 41.2% 29.4%
Family
Landesbank
Baden-
Wü embe g
100%
Boa d membe
CEO
Figu e A.2: Owne ship and go e nance s uc u e, G ei enbe ge AG, 2013
49
Table A.7: Sha eholde 1, G ei enbe ge AG, 2013
Va iable Coding Explana ion Sou ce
sh_name1 G ei enbe ge
Holding GmbH
Hoppens ed : Included in sha e-
holde da a
sh_sha e1 0.5887 Hoppens ed : Included in sha e-
holde da a
sh_ ype1 2 Holding company Ex e nal sou ce: „Gegens and
des Un e nehmens is die Ve -
wal ung eigenen Ve mögens und
das Hal en on Be eiligungen
an ande en Gesellscha en.“
h ps://www.no hda a.de/
G ei enbe ge +Holding+GmbH,
+Thu nau/Am sge ich +Bay eu h
+HRB+3604
sh_ am1 1 Owned by G ei en-
be ge amily which
is he sha eholde
wi h he (cumula-
i e) la ges (indi ec )
sha e.
Ex e nal sou ce: „He Heinz
G ei enbe ge häl 41,2 % de
An eile an de G ei enbe ge
Holding GmbH, wäh end de
allein o s and de G ei en-
be ge AG, He S e an G ei -
enbe ge , 29,4 % de An eile an
de G ei enbe ge Holding Gmbh
häl .“ h ps://g ei enbe ge .de/wp-
con en /uploads/2021/12/
G ei enbe ge _GB_2013.pd
sh_ ound1 0 No a na u al pe son
sh_ceo1 0 No a na u al pe son
sh_mgm 0 No a na u al pe son
sh_boa d1 0 No a na u al pe son
50
Table A.8: Sha eholde 2, G ei enbe ge AG, 2013
Va iable Coding Explana ion Sou ce
sh_name2 Baden-
Wü embe gische
Ve so gungsans al
ü Ä z e
Hoppens ed : Included in
sha eholde da a
sh_sha e2 0.062 Hoppens ed : Included in
sha eholde da a
sh_ ype2 6 Ins i u ional in es o bu
con olled by s a e en i y
Hoppens ed : An eil-
seigne : LBBW Asse
Managemen In es men -
gesellscha mbH, S u ga ,
100%; An eilseigne :
Landesbank Baden-
Wü embe g, S u ga ,
100%
Ex e nal sou ce: “Landes-
bank Baden-Wü embe g
[...] is a uni e sal bank
and he Landesbank
o some Fede al S a es
o Ge many [...] i is
Ge many’s bigges s a e-
backed landesbank lende .”
h ps://en.wikipedia.o g/wiki/
Landesbank_Baden-
W%C3%BC embe g
sh_ am2 0 No associa ed wi h G ei -
enbe ge amily
sh_ ound2 0 No a na u al pe son
sh_ceo2 0 No a na u al pe son
sh_mgm 2 0 No a na u al pe son
sh_boa d2 0 No a na u al pe son
51

Table A.9: Sha eholde 3, G ei enbe ge AG, 2013
Va iable Coding Explana ion Sou ce
sh_name3 Heinz G ei en-
be ge
Added as sha eholde be-
cause pa o he boa d and
holds sha es indi ec ly
Hoppens ed : Name included
in supe iso y boa d da a
sh_sha e3 0 No sha es held di ec ly; indi-
ec sha e included in sha e-
holde 1.
Hoppens ed : No included in
sha eholde da a
sh_ ype3 1 Na u al Pe son
sh_ am3 1 Membe o G ei enbe ge
amily which is he la ges
(cumula i e) sha eholde .
Ex e nal sou ce: „Fi -
meng ünde Heinz
G ei enbe ge “
h ps://www.ku ie .de/inhal .
s e an-g ei enbe ge -pack -
aus.5bad6 68-e0ec-420b-
bccc-7e3723346 41.h ml
sh_ ound3 1 Founde o he company Ex e nal sou ce: „Fi -
meng ünde Heinz
G ei enbe ge “
h ps://www.ku ie .de/inhal .
s e an-g ei enbe ge -pack -
aus.5bad6 68-e0ec-420b-
bccc-7e3723346 41.h ml
sh_ceo3 0 No a membe o manage-
men boa d
Hoppens ed : Managemen
boa d da a
sh_mgm 3 0 No a membe o manage-
men boa d
Hoppens ed : Managemen
boa d da a
sh_boa d3 1 Membe o supe iso y
boa d
Hoppens ed : Supe iso y
boa d da a
52
Table A.10: Sha eholde 4, G ei enbe ge AG, 2013
Va iable Coding Explana ion Sou ce
sh_name4 S e an G ei en-
be ge
Added as sha eholde be-
cause pa o he boa d and
holds sha es indi ec ly
Hoppens ed : Name included
in managemen boa d da a
sh_sha e1 0 No sha es held di ec ly; indi-
ec sha e included in sha e-
holde 1.
Hoppens ed : No included in
sha eholde da a
sh_ ype4 1 Na u al Pe son
sh_ am4 1 Membe o G ei enbe ge
amily which is he la ges
(cumula i e) sha eholde .
Ex e nal sou ce: „S e-
an G ei enbe ge [...]
Sohn des Fi meng ünde s
Heinz G ei enbe ge “
h ps://www.ku ie .de/inhal .
s e an-g ei enbe ge -pack -
aus.5bad6 68-e0ec-420b-
bccc-7e3723346 41.h ml
sh_ ound4 2 Rela i e o ounde Heinz
G ei enbe ge
Ex e nal sou ce: „S e-
an G ei enbe ge [...]
Sohn des Fi meng ünde s
Heinz G ei enbe ge “
h ps://www.ku ie .de/inhal .
s e an-g ei enbe ge -pack -
aus.5bad6 68-e0ec-420b-
bccc-7e3723346 41.h ml
sh_ceo4 1 CEO („Vo s ands o si zen-
de “)
Hoppens ed : Managemen
boa d da a
sh_mgm 4 1 Membe o managemen
boa d
Hoppens ed : Managemen
boa d da a
sh_boa d4 0 No a membe o supe iso y
boa d
Hoppens ed : Supe iso y
boa d da a
53
Table A.11: Family i m de ini ions, G ei enbe ge AG, 2013
De ini ion G ei enbe ge AG, 2013
Family sha e Cumula i e con olling sha e
held by membe s o he amily
58.87%
Owne ship Family sha eholde s hold mo e
han 25%
Yes
Owne ship & managemen
boa d
Mo e han 25% and a leas one
amily membe in managemen
boa d
Yes
Owne ship & supe iso y
boa d
Mo e han 25% and a leas one
amily membe in supe iso y
boa d
Yes
Owne ship & managemen
& supe iso y boa d
Mo e han 25% and amily mem-
be s in bo h boa ds
Yes
Owne ship & ounde CEO Mo e han 25% and CEO is he
ounde
No
Owne ship & o he amily
membe CEO
Mo e han 25% and CEO is an-
o he amily membe
Yes
Owne ship & ex e nal CEO Mo e han 25% and CEO is no a
amily membe
No
54
B Employmen and sales shocks
We ask whe he labo hoa ding may be mo e p onounced in amily i ms han in non- amily i ms
in he ace o sales shocks. In his Appendix, we p esen a simple eg ession analysis o es ima e
he sensi i i y o i m employmen o sales shocks. Speci ically, we es whe he his sensi i -
i y a ies be ween amily and non- amily i ms. The baseline speci ica ion o ou employmen
eg ession is:
log ei =β1log si +β2log si Fi +β3log si log agei +αi+δ +εi (10)
whe e ei is employmen and si is sales o i m iin yea ,Fi is a amily i m dummy a iable,
agei is he i m’s age, αiis a i m ixed e ec and δ is a yea ixed e ec o con ol o agg ega e
shocks. The coe icien β1measu es he sensi i i y o employmen o sales shocks in non- amily
i ms, and β2measu es he di e ence in ha sensi i i y be ween amily and non- amily i ms. In
a a ian o his speci ica ion, we also include in e ac ion e ms be ween yea ixed e ec s and he
amily i m dummy a iable o he i m’s age, o allow o he ac ha i ms may a y in hei
esponse o agg ega e shocks.
Table B.1 shows he eg ession esul s. While columns (1) and (2) show he esul s o es ima-
ions using ou baseline de ini ion o amily i ms, columns (3) and (4) show es ima ions including
dummies o amily ep esen a ion in he managemen and supe iso y boa ds. Las ly, columns
(5) and (6) shows esul s o he h ee subg oups o amily i ms led by he CEO, a amily membe
o an ex e nal CEO. The speci ica ions in columns (2), (4), and (6) include in e ac ion e ms be-
ween yea ixed e ec s and he espec i e amily i m dummy o he i m’s age. Ac oss all models,
we ind a simila , highly signi ican , employmen eac ion o sales shocks, indica ing ha i ms in
ou sample end o align hei wo k o ce numbe s wi h sales luc ua ions. Howe e , he es ima ed
e ec o amily i ms is negligible and lacks signi icance ac oss all speci ica ions, sugges ing ha
he esponse o amily i ms o sales shocks does no di e om ha o non- amily i ms.
55