Internationalisation as a boost for many firms: evidence from Germany

F ey, Raine ; Goldbach, S e an
A icle — Published Ve sion
In e na ionalisa ion as a boos o many i ms: e idence
om Ge many
Re iew o Wo ld Economics
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: F ey, Raine ; Goldbach, S e an (2024) : In e na ionalisa ion as a boos o many
i ms: e idence om Ge many, Re iew o Wo ld Economics, ISSN 1610-2886, Sp inge , Be lin,
Heidelbe g, Vol. 161, Iss. 3, pp. 911-963,
h ps://doi.o g/10.1007/s10290-024-00567-z
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/330753
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/
Vol.:(0123456789)
Re iew o Wo ld Economics (2025) 161:911–963
h ps://doi.o g/10.1007/s10290-024-00567-z
ORIGINAL PAPER
In e na ionalisa ion asaboos o many i ms: e idence
omGe many
Raine F ey1· S e anGoldbach1
Accep ed: 1 Oc obe 2024 / Published online: 2 Decembe 2024
© The Au ho (s) 2024
Abs ac
The ad an ages o globalisa ion ha e been inc easingly called in o ques ion, and
p o ec ionis endencies ha e en e ed he s age. So wha expe iences ha e i ms had
a e going in e na ional in an open economy o e he pas yea s? Wi h espec o
o eign in es o akeo e s o ini ially na ional i ms, we see p edominan ly posi i e
e ec s o acqui ed manu ac u ing i ms in e ms o p oduc i i y, sales and expendi-
u es on he labou o ce – likely due o highe employmen – in he sho and long
e m. Looking a he esul s, i m size ma e s: posi i e e ec s a e s onge among
la ge i ms. In he case o i ms s a ing o in es ab oad, posi i e e ec s a e e y
a e and limi ed o sho - e m sales o acqui ing small manu ac u ing i ms. All in
all, he la gely posi i e e idence is gene ally suppo i e o in e na ionalisa ion and
hus la gely con adic s he nega i e iews some imes p esen in he public sphe e,
hough e en una ec ed i ms may see hemsel es as ela i e lose s.
Keywo ds Globalisa ion· Fi m acquisi ion· M&A· P oduc i i y· Sales· In angible
asse s· Knowledge· Technology· Labou cos s· Employmen · Wages· Fi m
he e ogenei y
JEL Classi ica ion D22· D24· F23· G34
1 In oduc ion
In he pas , globalisa ion was la gely seen as inc easing he wel a e o all coun-
ies in ol ed. Howe e , mo e ecen ly, some go e nmen s ha e shi ed he ocus
o po en ial nega i e e ec s. In his con ex , mo e emphasis is placed on issues
* Raine F ey
aine [email p o ec ed]
S e an Goldbach
s e [email p o ec ed]
1 Di ec o a e Gene al Economics, Deu sche Bundesbank, Mainze Lands asse 46,
60325F ank u AmMain, Ge many
912
R.F ey, S.Goldbach
such as c oss-bo de echnology ans e o he de imen o he a ge i m,1 local
di es men and labou layo s, o public secu i y isk h ough new o eign owne s
in me ge s and acquisi ions (M&A) ansac ions. These conside a ions played a
ole in Eu opean coun ies in oducing new laws ha enable go e nmen s o o -
bid o eign akeo e s (see Eu opean Union (2019)). Simila ly, coun ies such as he
Uni ed S a es ha e in oduced ade es ic ions, ci ing po en ially un ai compe i-
ion while a ac ing o eign in es men h ough subsidies (In la ion Reduc ion Ac ).
Con e sely, howe e , limi a ions imposed on o eign acqui e s can lead o o eign
go e nmen s in oducing coun e measu es a ge ing domes ic in es men ab oad.
Wi h na ional es ic ions, i ms isk he po en ially subs an ial p o i s om o m-
ing mul ina ionals o being pa o a mul ina ional g oup. While in he case o e i-
cal o eign di ec in es men (FDI), i ms dis ibu e he p oduc ion p ocess ac oss
se e al coun ies o p o i om an in e na ional di ision o labou ,2 ho izon al FDI
is d i en by he de elopmen o new ma ke s o a i m’s p oduc s.3 By engaging in
FDI, i ms especially in end o inc ease hei e iciency. Add essing he e ec s o
i ms’ owne ship changes on economic e iciency, as one ea ly example o ample
s udies in he manu ac u ing sec o , Lich enbe g and Siegel (1987) analyse he
s ong inc ease in M&A in he Uni ed S a es back in he 1970s and 1980s, wi h sig-
ni ican p oduc i i y gains in he yea s a e owne ship change.4
So does globalisa ion only ha e posi i e e ec s while es ic ions do ha m o
i ms? To answe his ques ion, we need o ake a mo e de ailed look a di e en
i m ypes and a ious economic indica o s. In pa icula , ou s udy add esses po en-
ial in e na ionalisa ion e ec s o he e ogeneous i ms wi h espec o p oduc i i y,
sales, andin angible asse s. The s ock o in angible asse s ep esen s di e en kinds
o knowledge and b and epu a ion and is likely accompanied by posi i e longe -
e m p oduc i i y and compe i i eness implica ions.5 Finally, we use labou cos s
1 Ge s enbe ge (2018) add esses his discussion wi h espec o Ge man companies acqui ed by Chinese
i ms. Mo e gene ally, wi h espec o he pe o mance o i ms acqui ed by Chinese in es o s, see Fues
e al. (2021).
2 Fo he dis inc ion be ween e ical and ho izon al FDI, see Helpman (1984) and Ma kusen (1984).
3 Fa ell and Shapi o (1990) show ha ho izon al me ge s can inc ease wel a e – bu in combina ion
wi h signi ican inc eases in ma ke concen a ion, nega i e e ec s may a ise due o highe sales p ices.
Ca es (1974) sees FDI as a sou ce o alloca i e and echnical e iciency o he hos coun y.
4 In addi ion, acco ding o Helpman e al. (2004), he mos p oduc i e i ms engage in o eign ac i i ies.
5 In angible asse s include componen s such as so wa e, in ellec ual p ope y, b ands and inno a i e
business p ocesses. Thei cha ac e is ics acili a e p o i ing om economies o scale, and legal i les like
pa en s and copy igh s make i possible o secu e compe i i e ad an ages is-à- is compe i o s. This may
esul in highe p oduc i i y, ma ke powe and ma ke concen a ion (see C ouze and Ebe ly (2019)).
Da ies and Ma kusen (2020) also epo posi i e e ec s and obse e ha mul ina ionals p o ide hei
o eign a ilia es wi h knowledge-based asse s. Thus, a e in es ing ab oad, a echnology ans e wi hin
he newly o med mul ina ional may ake place (see Teece (1977)) and he e lies an especially s a egic
mo i e o expanding in angible asse s.
913
In e na ionalisa ion asaboos o many i ms: e idence om…
o p oxy employmen and wages – knowing ha his is a a he c ude measu e.6,7
The li e a u e documen s he gene al ele ance o hese a iables: ollowing a o -
eign acquisi ion, Guadalupe e  al. (2012) ind a subs an ial inc ease in sales and
p oduc i i y and a g ea e likelihood o inno a ing o he acqui ed i ms.8 In S ie-
bale and Vencappa (2018), he p oduc i i y e ec seems especially la ge o small
i ms. A nold and Ja o cik (2009) ind ha a ge i ms ha e highe p oduc i i y as a
esul o es uc u ing whe e acqui ed plan s inc eased in es men , employmen and
wages.9 Ai ken e al. (1996) es ima e posi i e wage e ec s o o eign-owned i ms
in he Uni ed S a es, Mexico and Venezuela. Wi h espec o knowledge ans e /
inno a ion, Ja o cik and Poelhekke (2017) p o ide empi ical suppo o con inuous
injec ions o headqua e se ices in o o eign plan s.
We add subs an ially o his li e a u e by add essing i m he e ogenei y. In doing
his, he analysis may de ec i m ypes ha p o i bu also hose ha do no p o i o
a e e en nega i ely a ec ed, such as in he case o labou o ce layo s o shi s o
in angible asse s wi hin he i m conglome a e ac oss bo de s (“ echnology he ”)
– e en s ha may b ing wi h hem nega i e a i udes owa ds globalisa ion. Fi s ,
o ou knowledge, ou s udy is unique as we sepa a ely add ess he manu ac u ing
and he se ices sec o s: p e ious s udies we e mos ly limi ed o he manu ac u -
ing sec o , e.g. Guadalupe e al. (2012) o Spanish i ms, Bi can (2019) o Tu kish
plan s and S iebale and Vencappa (2018) o Indian i ms, o hey co e ed all sec o s
wi hin one sample (e.g. Hijzen e al. (2013) o se e al coun ies). We see a sepa-
a e analysis o he se ices sec o as app op ia e, as he se ices sec o has apidly
gained in e na ional economic ele ance o e he las ew decades.
Second, we in es iga e he sho and long- e m e ec s o in e na ionalisa ion on
i m pe o mance o unde s and he dynamics a e a i m acquisi ion. Depending on
he pe o mance a iable, e ec s a e expec ed in he sho e m al eady (e.g. sales
may inc ease a he quickly) o mo e in he long e m (e.g. p oduc i i y, in angible
asse s o employmen /wages). We de ine ou sho - e m ho izon as he ime span
wi hin he yea o he acquisi ion. Meanwhile, ou long e m comp ises he i s h ee
yea s a e he acquisi ion.10 In he case o ou a iable in angible asse s – whe e
an inc ease may ul ima ely also esul in highe p oduc i i y, compe i i eness and
6 Ou da ase con ains a mo e da a on labou cos s han, o example, employmen . Thus, in ou
app oach, we use o al labou cos s as a combined measu e o employmen and wages. The easons o an
inc ease in labou cos s may be highe wages o mo e employmen , ce e is pa ibus.
7 The e may also be some in e dependencies be ween hese a iables: Conyon e al. (2002) ind wage
inc eases due o highe p oduc i i y. Egge and K eickemeie (2013) a gue ha in e na ional companies
a e mo e compe i i e and a e hus able o pay highe wages. Koch and Smolka (2019) p o ide e idence
ha he acqui ed i ms hi e highly skilled wo ke s and p o ide wo ke aining, which esul s in highe
compe i i eness.
8 Complemen a ily, Be and and Zi ouna (2008) obse e an inc ease in p oduc i i y bu no in p o i s
a e a me ge . They conclude ha i ms may edis ibu e e iciency gains wi hin he i m conglome a e.
9 Fo es ablishmen s in he Uni ed S a es, Da is e al. (2014) show ha p i a e equi y buyou s may yield
posi i e e ec s o some plan s bu nega i e labou e ec s o o he plan s, he eby imp o ing he ope a -
ing ma gin ma e ially.
10 Egge e al. (2020) ind ha pos -acquisi ion wage e ec s ake a o al o ou yea s o de elop and
emain cons an he ea e . We look only a he i s h ee yea s a e he acquisi ion, as a u he ex en-
sion o he ime ho izon comes wi h a sizeable dec ease in obse a ions.
914
R.F ey, S.Goldbach
ma ke powe – we po en ially cap u e u he posi i e e ec s ha a e el e en
beyond ou h ee-yea ime ho izon.
Thi d, we allow o po en ial di e ences in i m size like, o example, S iebale
and Vencappa (2018), who documen posi i e p oduc i i y e ec s concen a ed on
small ( a ge ) i ms.11 Wi h espec o la ge me ge s, Gugle e al. (2003) ind ha
p o i s may also inc ease h ough highe ma ke powe o e a longe ime ho izon.
Howe e , we canno conside he s uc u al implica ions a ising om changes in he
ma ke s uc u e wi hin ou app oach.12
Finally, we con ibu e o he la ge empi ical li e a u e ha deals wi h he e ec
o ou wa d FDI on domes ic economic ac i i y. Desai e  al. (2009) ind o US
manu ac u ing i ms ha ou wa d FDI inc eases domes ic in es men s and wages.
Goldbach e al. (2019) con i m he indings o in es men s o Ge man i ms using
he Mic oda abase Di ec in es men (MiDi) compiled by he Deu sche Bundes-
bank. Howe e , hey only ocus on ou wa d FDI, while we examine he e ec s o
in e na ionalisa ion in bo h di ec ions. O he empi ical pape s also use he MiDi,
e.g. Becke and Mündle (2010) examine po en ial in a-g oup labou subs i u ion
o Ge man manu ac u ing mul ina ional i ms. Ou analysis de ia es om hei
app oach because we concen a e on Ge man i ms engaging in FDI o he i s
ime. By con as , Becke and Mündle (2010) mos ly ocus on mul ina ional i ms,
which ha e al eady o eign a ilia es. Ano he example is Con educa and Kazako a
(2018). They e en use a a iable (“em1” = eason o ini ial epo ) om he MiDi
ha epo s he en y mode. Howe e , his in o ma ion is op ional (no manda o y)
and o en no p o ided. Tha is why we decided o use an al e na i e iden i ica ion
s a egy ha elies on ma ching be ween he Deu sche Bundesbank’s balance shee
da abase o Ge man i ms (JANIS) and MiDi.13
When we add ess bo h inwa d and ou wa d FDI wi hin one se ing a he mic o
le el, we use da a om Ge many, a la ge open economy. A he end o 2019, he
consolida ed FDI s ocks o Ge man mul ina ionals ab oad amoun ed o €1.4 il-
lion. The FDI holdings o o eign pa en companies in a ilia es loca ed in Ge many
s ood a €0.6 illion and a e hus also impo an om an economic pe spec i e. The
exis ing li e a u e mos ly in es iga es he e ec s on i ms s emming om an owne -
ship change o a o eign in es o , e.g. Ha ison and Lipsey (1996), Guadalupe e al.
(2012), while Ash a e al. (2016) ocus on i m p oduc i i y and Be and e  al.
(2012) concen a e on esea ch and de elopmen . One excep ion is Szücs (2014),
who add esses he in ol emen o bo h M&A acqui e s and a ge s and inds ha
he a ge s’ R&D dec eases and he acqui e s’ R&D- o-sales in ensi y d ops due o
11 De ia ing om hei app oach, we use he sum o angible and in angible asse s as a size measu e
ins ead o “sales”. We expec his o be mo e app op ia e o ou s udy, which deals wi h di e en sec o s.
12 I la ge i ms s eng hen an al eady s ong posi ion, i may ha e implica ions o o e all compe i ion in
a sec o . The p ospe i y o he economy as a whole does no necessa ily inc ease. I small i ms a e busy
ca ching up, i ini ially migh no ha e he same implica ions o compe i o s. In addi ion, in he long
e m, i ms can also p o i om echnological clus e s ab oad h ough local spillo e s (see Ja e e al.
(1993)).
13 We conduc ed a obus ness check wi h an al e na i e iden i ica ion s a egy ha elies on he mode o
en y. Since his in o ma ion is o en no p o ided, we do no ha e enough obse a ions in he ea men
g oup o a meaning ul in e p e a ion o hese esul s.

915
In e na ionalisa ion asaboos o many i ms: e idence om…
s ongly inc easing sales. Howe e , he does no accoun o he e ogenei y.14 Thus,
we add o he ex emely small numbe o s udies ha also look a domes ic i ms’
po en ial oppo uni ies h ough going global.15 This is impo an as, o gi e an
example, in ligh o na ional p o ec ionis endencies wi h espec o c oss-bo de
akeo e s, ecip ocal ac ions by o he coun ies could, in u n, jeopa dise domes ic
i ms’ in es men oppo uni ies ab oad.
Fo ou analysis, we ma ch Ge man mic o da a o FDI wi h i m balance shee
da a. These da ase s a e p o ided by he Deu sche Bundesbank. This allows us o
ely on o icial alues p o ided by i ms. In addi ion, he da a also allows o clea
iden i ica ion o pu ely na ional i ms and o i ms ha go in e na ional. Fo in es -
men s unde aken by Ge man i ms ab oad, we use o al ou wa d FDI, which
includes M&A as well as g een ield in es men .16 As bo h ypes o in es men may
esul in he in e na ionalisa ion o ini ially domes ic i ms, we hus ob ain a com-
ple e pic u e o he associa ed in e na ionalisa ion e ec s, al hough we canno dis in-
guish be ween he con ibu ions o he wo ypes.17
To accoun o a po en ial selec ion bias in ou empi ical analysis, like many
au ho s, e.g. Guadalupe e  al. (2012) and S iebale and Vencappa (2018), we i s
add ess he acquisi ion decision and hen conduc p opensi y sco e ma ching o es i-
ma e he likelihoods o i ms being chosen o acquisi ion. In doing his, Guadalupe
e al. (2012) ind o Spanish manu ac u ing i ms ha acquisi ions a e ocused on
he mos p oduc i e i ms wi hin indus ies in he acqui e s’ in e es o inc easing
inno a ion and p oduc i i y. This wo-s age p ocedu e allows o weigh ed dynamic
di e ence-in-di e ences (DiD) es ima ions o pos -acquisi ion e ec s.18
All in all, ou ou come o in e na ionalisa ion e ec s shows ha in he a e ma h
o an acquisi ion in he manu ac u ing sec o , i ms expe ience posi i e e ec s on
p oduc i i y, sales and expendi u e on he labou o ce in he sho and long e m
– wi h he long- e m e ec s gene ally being s onge o la ge i ms. In he se ices
14 Ano he s and o li e a u e looks a e ec s in he hos and home coun ies o FDI, e.g. Ha ms and
Méon (2018). Chen (2011) inds highe p oduc i i y gains, sales and employmen only when he o eign
pa en s come om indus ial coun ies a he han de eloping coun ies. Hijzen e al. (2013) show ha
wage e ec s a e la ge in de eloping coun ies.
15 Wi h espec o Ge man in es men ab oad, we only add ess Ge man i ms in es ing o he i s ime
in o he coun ies. In addi ion, he pa en companies o al eady exis ing mul ina ionals may also p o i
om en e ing u he o eign ma ke s. In gene al, i only some coun ies in oduce es ic ions, in es o s
may ci cum en hese policies ia o he des ina ion economies.
16 Though he Bundesbank da a o m pe mi s use s o di e en ia e be ween g een ield in es men and
M&A, his op ional ield is usually le emp y. S udies elying on hese da a a e e y a e, e.g. Bialek and
Weichen iede (2021) wi h a ocus on FDI egula ion o Con educa and Kazako a (2018).
17 Nocke and Yeaple (2007) emphasise he ele ance o i ms’ choice o in es men ype. Henna and
Pa k (1993) ind ha Japanese in es o s use M&A in he case o ela i ely weak compe i i e ad an ages;
o he wise g een ield in es men is a mo e e icien way o ans e ad an ages. Da ies e al. (2018) ind
g een ield in es men mo e elian on i ms’ own capaci ies, on he o igin coun y’s compa a i e ad an-
ages and he des ina ion coun y’s axes.
18 Bals ik and Halle (2010) u he mo e conclude ha o eign i ms pick la ge, high-wage and high-
p oduc i i y i ms, which is also la gely co obo a ed by he li e a u e. In addi ion, F ey and Hussinge
(2011) show ha a ge i ms a e delibe a ely chosen o inc ease he echnological compe ency o he
g oup as a whole. C ouze and Ebe ly (2019) e en see a shi in ele ance om physical o in angible
asse s in he Uni ed S a es o e ime.
916
R.F ey, S.Goldbach
sec o , he majo e ec s b oadly disappea . The sales e ec is limi ed o he sho
e m. Fo in angible asse s, he posi i e/nega i e e ec depends on i m size. In case
o i ms ha s a in es ing ab oad, posi i e e ec s a e e y a e and limi ed o he
sho - e m sales o small manu ac u ing i ms. Again, we ind no e idence ha i ms
es uc u e a he expense o he labou o ce a home. Thus, ou esul s unde line
ha c oss-bo de in es men s o en s eng hen i ms and make hem mo e compe i-
i e – hough he e a e also nume ous i ms ha a e le behind and, in e y a e
cases, nega i ely a ec ed. Addi ionally, pa icipa ion in in e na ional business is no
andom: la ge i ms a e mo e likely o be chosen as a ge s.
The emainde o he pape is s uc u ed as ollows. Sec ion 2 desc ibes ou
da a, ollowed by he empi ical me hodology in Sec .3. The esul s a e p esen ed in
Sec .4. We conduc se e al obus ness checks in Sec .5. Sec ion6 b ie ly concludes.
2 Da a
To analyse he e ec o FDI on domes ic pe o mance, we use wo mic o da a-
se s, bo h p o ided by he Deu sche Bundesbank.19 In o ma ion on o eign in es-
o s’ ac i i ies in Ge man i ms (inwa d FDI) and on he o eign in es men ac i i-
ies o Ge man companies (ou wa d FDI) is ob ained om MiDi, a comp ehensi e
annual da abase o Ge man FDI posi ions. MiDi p o ides in o ma ion on balance
shee i ems, owne ship s uc u e and addi ional in o ma ion such as an economic
sec o classi ica ion o each Ge man a ilia e owned by a o eign pa en company
(and o eign a ilia es o Ge man pa en i ms). One pa icula ad an age o MiDi
is ha epo ing by Ge man i ms is manda o y unde Ge man ede al law. In o ma-
ion on domes ic pe o mance and on se e al o he pa en -le el a iables is aken
om JANIS.20 This da ase is an ex ension o USTAN, a da ase o co po a e bal-
ance shee s a is ics. The da a a e p ima ily ex ac ed om annual accoun s (bal-
ance shee , p o i and loss accoun s) and inancial s a emen s. We make use o i m-
speci ic in o ma ion on o al sales, alue added ( o es ima e p oduc i i y), in angible
asse s ( o p oxy i m ele an knowledge) and labou cos s o cap u e employmen
and wages.21 Appendix Table6 p o ides an o e iew o he de ini ions o ele an
i m a iables. Acco ding o he balance shee in o ma ion om JANIS, in angible
asse s a e he sum o he ollowing posi ions: “concessions, indus ial p ope y and
simila igh s and asse s as well as licences”, “goodwill”, “paymen s on accoun o
19 The mic o da a a e con iden ial and only accessible in anonymised o m a he headqua e s o he
Bundesbank in F ank u , Ge many.
20 The da a epo (h ps:// www. bunde sbank. de/ esou ce/ blob/ 901916/ 4a1c0 d2604 04103 24e8e 306a4
1a0b9 71/ mL/ 2023- 14- janis- da a. pd ) highligh s ha “JANIS consis s o indi idual inancial s a emen s
o non- inancial co po a ions which a e p o ided om se e al sou ces: inancial s a emen s ecei ed by
he Deu sche Bundesbank in he con ex o he c edi assessmen and om public sou ces like he Bun-
desanzeige .” Al hough he di e en da a sou ces a e consis en and a e subjec o quali y checks, he aw
da a exhibi an unbalanced panel s uc u e. We adjus he da a o compa e ou indings wi h he li e a u e,
which dis inguishes be ween sho and long- e m e ec s.
21 The i ms epo hei sales o o eign a ilia es ( om he MiDi da ase ) and he sales o hei domes ic
en i ies ( om he JANIS da ase ) independen ly.
917
In e na ionalisa ion asaboos o many i ms: e idence om…
in angible asse s” and “in e nally gene a ed indus ial igh s and simila igh s and
asse s”. “Labou cos s” a e he sum o he posi ions “sala ies and wages” and “social
secu i y and expendi u e o company pension unds and pensions paid”. We ma ch
JANIS wi h MiDi and keep ma ched obse a ions as well as unma ched obse a-
ions. Fi ms ha a e Ge man pa en companies and Ge man a ilia es o o eign
companies a he same ime a e excluded om he analysis in o de o make a clea
dis inc ion be ween he e ec s o inwa d and ou wa d in es men .22 In addi ion, we
only keep i m-yea obse a ions o i e yea s in a ow.23
In he i s analysis, we compa e Ge man companies ha a e aken o e (“ a -
ge s”) by o eign i ms wi h i ms ha emain domes ically owned. We only keep
hose i ms in he con ol g oup ha ha e no ma ch wi h any o eign a ilia e – no
ou wa d FDI – in he pe iod conside ed. The second analysis compa es Ge man
companies engaging in FDI o he i s ime (“acqui e s”)wi h i ms ha emain
pu ely domes ic.24 In u n, we ensu e ha ou compa ison g oup only consis s o
pu ely domes ic i ms ha ha e no ma ch wi h a o eign in es o in MiDi da a – no
inwa d FDI. All in all, we end up wi h an unbalanced panel o he ime pe iod om
1999 o 2018, wi h abou 1,900 Ge man i ms aken o e by o eign companies, 900
domes ic i ms going global o he i s ime, 57,000 pu ely na ional companies in
ou con ol g oup, and 360,000 (334,000) i m-yea obse a ions in o al o “ a ge ”
(“acqui e ”) i ms.25
Table1 p esen s da a abou he Ge man i ms in ol ed in akeo e s by o eign
in es o s in mo e de ail. In acco dance wi h he li e a u e, we concen a e on pos-
sible i m cha ac e is ics wi h a ime lag o one yea , which we conside ele an o
o eign in es o s. In he able, he panel on he le (“no a ge ”) p esen s obse a-
ions o he con ol g oup, while he panel on he igh (“as a ge ”) e e s o only
hose obse a ions (numbe o i ms) a he ime when a o eign in es o acqui es
a Ge man i m. We p o ide summa y s a is ics o all sec o s, he manu ac u ing
22 Looking a he o iginal MiDi da abase (1999 o 2018) wi hou any adjus men s, we obse e abou
18,000 epo ing uni s as Ge man pa en companies and abou 41,000 domes ic a ilia es o o eign
in es o s. Ano he app oxima ely 2,000 i ms a e Ge man pa en s and Ge man a ilia es (o a o eign
ul ima e in es o ) in he same yea .
23 Fo ou a ge s and acqui e s, we addi ionally equi e ha hey a e in ou da a o a leas wo yea s
be o e he akeo e and o h ee yea s a e wa ds. This is in line wi h he ela ed li e a u e. An analysis
o i ms ha closed wi hin his pe iod is no easible. This is because i is no possible o dis inguish
be ween i m closu es and missing da a. The es ic ions we impose do no a ec he composi ion o
indus ies o he egional dis ibu ion. We do see some sel -selec ion o i ms ha su i e o e ime and
become la ge on a e age (in e ms o o al asse s, o al sales and – when p o ided – employmen ). Fo
u he de ails, see Appendix Table8. We also checked whe he his imposed es ic ion a ec s ou base-
line esul s. Since he indings do no change much, we a gue ha ou chosen a iables in he i s s age
o he p opensi y sco e ma ching con ols o his po en ial sel -selec ion bias.
24 “Fi s ime” e e s o he change in he s a us wi hin he sample pe iod. Fi ms in es ing ab oad o
he i s ime do no necessa ily need o acqui e ano he o eign company. They can also es ablish a new
o eign a ilia e (“g een ield in es men ”). Ou da ase does no allow us o dis inguish be ween he wo
ypes o FDI. Ne e heless, we use he e m “acqui e s” h oughou he pape o he sake o simplici y.
25 The e is a h eshold o €3million o i ms’ FDI ilings. Thus, we only conside FDI o some sig-
ni icance. Howe e , he ela i e impo ance o he o eign in es men may di e depending on he pa en
companies’ size.
918
R.F ey, S.Goldbach
sec o and he se ices sec o .26 The sec o al dis inc ion is e ained o ou es ima-
ions. We a e awa e ha a clea assignmen is a he di icul , especially o la ge
companies as hey a e o en ac i e in bo h o hese a eas. Thus, we assume ha
he co e ac i i ies epo ed by he i ms a e ele an o hei o eign in es men s.
As ega ds Ge man i ms going ab oad o he i s ime, we a e dealing wi h ela-
i ely small i ms o which sec o al assignmen is gene ally clea e . Wi h espec o
i m cha ac e is ics, Table1 highligh s di e ences be ween he ea men and con ol
g oups – wi h he excep ion o p io o al ac o p oduc i i y. The same conclusion
can be de i ed om Table2, in which i ms going global o he i s ime a e com-
pa ed o companies ha emain pu ely domes ic. We accoun o hese di e ences in
ou es ima ion s a egy.
Table3 p esen s he loca ions o o eign in es o s ha acqui ed Ge man compa-
nies and he des ina ions o in es men s by Ge man i ms ha a e in es ing ab oad
o he i s ime. Again, we also dis inguish be ween he manu ac u ing and he
se ices sec o s. In e ms o bo h he absolu e numbe o i ms and he olume o
FDI, i ms om Wes e n Eu ope and No h Ame ica ep esen he mos impo an
in es o s in Ge man companies, ollowed by Asian in es o s. Ge man i ms in es -
ing ab oad concen a e p ima ily on Wes e n and Eas e n Eu opean coun ies.
This could ela e o c oss-bo de alue chains. Asia and No h Ame ica also play
an essen ial ole as des ina ions. In addi ion, we obse e a much highe olume o
“new” FDI in Ge many han ha o Ge man companies in es ing ab oad. To in e -
p e hese di e ences, we ha e o conside he likelihood ha a la ge numbe o o -
eign i ms al eady o m pa o a mul ina ional ha is adding a uni o i s po olio
o i ms. In con as , ou Ge man in es o s a e en e ing he in e na ional “playing
ield” o he i s ime.27 Thus, hese domes ic in es o s a e mos likely smalle on
a e age han he acquisi ions o o eign i ms in Ge many, which is e lec ed in he
FDI igu es.
3 Empi ical app oach
In ou empi ical se ing, we y o iden i y he causal e ec o a change om na ional
o o eign owne ship on he a ge i m’s pe o mance and o a i s in es men
ab oad on he pa en company’s pe o mance. To keep ou desc ip ion o he econo-
me ic p ocedu e sho bu s ill comp ehensi e, we explain ou app oach below by
solely looking a he scena io in which he o eign in es o buys a domes ic com-
pany; he p ocedu e o i s - ime FDI by a Ge man pa en company is he same.
Fo ou analysis, we ely on a DiD app oach ha we combine wi h p opensi y sco e
ma ching echniques.
26 The manu ac u ing sec o comp ises he wo-digi NACE Re . 2 codes 10 o 35. The se ices sec o
consis s o he codes 45 o 63, 68 and 69, 71 o 82, and 85 o 96. Thus, we exclude he inance sec o ,
holdings, households and o ganisa ions as hey a e expec ed o ac di e en ly compa ed o he es o he
se ices sec o .
27 In ou s udy, we do no conside al eady exis ing Ge man mul ina ionals expanding hei in e na ional
in es men .
925
In e na ionalisa ion asaboos o many i ms: e idence om…
whe e
yi
s ands o ou pe o mance a iables o in e es : ac o p oduc i i y, sales,
in angible asse s and labou cos s o i m i in pe iod – all in loga i hms39; he
dummy
Fis
equals 1 in he case o an acquisi ion and 0 o he wise;
𝜇i
and
𝜌j
a e
i m and indus y-yea ixed e ec s; and
𝛼
is he cons an . Due o he inclusion o
i m-le el ixed e ec s, he coe icien
𝛽
add esses he e ec h ough he change in
owne ship – no owne ship i sel . To accoun o se ial co ela ion, we clus e a he
i m le el.
The es ima ed e ec s may di e no only wi h espec o sec o and ime bu also
wi h espec o i m size:
(4)
y
i =𝛼+
2
∑
k=0
𝛽k∗Fi −k+
2
∑
k=0
𝛾k∗smalli∗Fi −k+𝜇i+𝜌j +𝜀
i
Table 4 P obabili y o being a ge and being acqui e
P obi es ima ion. The dependen a iable is he bina y indica o “change” in pe iod . The uni o obse -
a ion is i m-yea obse a ion. The sample pe iod is 2001 o 2018. Sec o ixed e ec s and ime-speci ic
ixed e ec s a e included bu no epo ed. Robus s anda d e o s (clus e ed by i m) a e in pa en heses.
***, ** and * deno e signi icance a he 1%, 5% and 10% le el, espec i ely
Ta ge Acqui e
Manu ac u ing Se ices Manu ac u ing Se ices
(1) (2) (3) (4)
Log o al asse si -1 0.129*** 0.195*** 0.279*** 0.245***
(0.028) (0.019) (0.032) (0.025)
Log in angible asse si -1 0.026*** 0.008 0.035*** 0.053***
(0.008) (0.007) (0.009) (0.011)
Log TFPi -1 −0.309* 0.064 −0.896*** 0.029
(0.179) (0.061) (0.164) (0.098)
Log labou cos si -1 0.186 0.040 0.833*** 0.039
(0.156) (0.054) (0.145) (0.089)
Log salesi -1 0.079*** 0.010 0.002 −0.004
(0.030) (0.017) (0.033) (0.020)
Re u n on equi yi -1 −0.000 −0.001** −0.000 −0.000***
(0.000) (0.000) (0.000) (0.000)
Log ixed asse si -1 0.028 −0.132*** 0.025 −0.104***
(0.018) (0.010) (0.017) (0.015)
TFP g ow hi -1 0.066 0.077 −0.021 0.143**
(0.065) (0.047) (0.090) (0.071)
Obse a ions 107,959 199,498 96,459 181,567
39 P opensi y sco e ma ching is an app oach o he c oss-sec ion. Since we es ima e ou e ec s wi hin
a panel da ase , a i m can exhibi a posi i e weigh in one yea and a missing alue in ano he yea . We
ca y he p opensi y sco e o a i m o wa d i he e is a missing alue un il he p opensi y sco e ma ch-
ing es ima es a new weigh o his i m.

926
R.F ey, S.Goldbach
whe e
smalli
equals 1 o small i ms below o equal o he median o he sum
o angible and in angible asse s, and 0 o he wise. We classi y he “ ea ed” i ms
as small o la ge based on hei s a us one yea be o e he ea men . Since adius
ma ching allows o mul iple con ol obse a ions, he s a us o a i m may change
o e ime. To deal wi h hese di e ences, we coun he numbe o small and la ge
s a uses o he i ms in he con ol g oup ( i ms wi h a posi i e weigh om he
p opensi y sco e ma ching in a speci ic yea ) o e he comple e ime pe iod. I he
majo i y is small (la ge), we de ine he i m in he con ol g oup as small (la ge).
This de ini ion keeps he s a us o he ea men and con ol g oups cons an o e
ime and we do no ha e o d op obse a ions.
4 Resul s
In e na ionalisa ion is like a shock o a i m and may ha e an impac on i s economic
de elopmen . Figu e 1a p esen s he e en s udy’s coe icien plo s o o al ac o
p oduc i i y, sales, in angible asse s and labou cos s o all i ms.40 The uppe
(lowe ) ow shows he esul s o a ge s (acqui e s). Fu he mo e, we dis inguish
be ween he manu ac u ing (le panel) and he se ices ( igh panel) sec o s. The
e ec s o in e na ionalisa ion in he e en s udy a e plo ed o he yea o he e en
and he wo subsequen yea s. To check he alidi y o he CTA, we addi ionally
plo he wo yea s p eceding he e en . The insigni ican coe icien s in pe iods -2
and -1 o Fig. 1a sugges ha he CTA is no iola ed, wi h only one excep ion
(in angible asse s o acqui e s in he se ices sec o ). We ind no signi ican e ec
in he pe iods , + 1 and + 2 o acqui e s (lowe ow), nei he in he manu ac u -
ing no he se ices sec o o all ou economic a iables. Fu he mo e, he se ice
sec o ( igh panel) is, in gene al, no a ec ed by he ea men . Fo “ a ge ” i ms in
manu ac u ing sec o , howe e , we ind some signi ican posi i e ea men e ec s,
which a e lis ed in panel (a) o Table5. The inc emen al e ec s diminish o e ime.
We es ima e a signi ican ea men e ec o p oduc i i y o 5.1% in , 5.0% in + 1
and 3.2% in + 2 (sum o all coe icien s: 13.3%). The e ec on sales is la ge : 7.7%
in , 7.2% in + 1 and 6.5% in + 2 (sum o all coe icien s: 21.4%). Finally, being
“ a ge s” leads o highe labou cos s o 5.8% in , 5.9% in + 1 and 3.4% in + 2
(sum o all coe icien s: 15.1%). Fo in angible asse s, he signi ican e ec is only
exis en in (16.6%) and + 1 (17.4%) bu economically la ge (sum o bo h coe i-
cien s: 33.9%). The e o e, we see ha Ge man a ilia es bene i om being a “ a ge ”
by o eign in es o s.
Fig. 1 a: E en s udy o all i ms ( i s ow: a ge s; second ow: acqui e s; le : manu ac u ing sec o ;
igh : se ices sec o ). b: E en s udy o small i ms ( i s ow: a ge s; second ow: acqui e s; le : man-
u ac u ing sec o ; igh : se ices sec o ). c: E en s udy o la ge i ms ( i s ow: a ge s; second ow:
acqui e s; le : manu ac u ing sec o ; igh : se ices sec o )
▸
40 The e en s udy app oach es ima es Eq.(3) bu s a s a k = −2 ins ead o k = 0. Fu he mo e, ins ead
o using Eq.(4) o dis inguish be ween small and la ge i ms, we use Eq.(3) o small and la ge i ms as
subsamples.
927
In e na ionalisa ion asaboos o many i ms: e idence om…
928
R.F ey, S.Goldbach
Figu e1b highligh s he e en s udy g aphs o small i ms. Since he esul s a e
simila o all i ms, we concen a e on he di e ences be ween bo h samples. Again,
mos coe icien s in -2 and -1 a e insigni ican wi h only one excep ion (in angi-
ble asse s o a ge s in he se ices sec o ). Al hough, we ind again only signi i-
can esul s in he manu ac u ing sec o o “ a ge ed” i ms, hese e ec s endu e
sho e and a e quan i a i ely smalle . Only sales a e signi ican in he pe iods , + 1
and + 2. All o he a iables become insigni ican in he las pe iod compa ed o all
i ms. Since almos all coe icien s a e smalle , wi h he only excep ion o in angi-
ble asse s in pe iod , we ind weake e ec s o small i ms compa ed o all i ms.
Table5 panel b p esen s he indi idual sho - e m and long- e m esul s o all ou
economic a iables o small i ms.
Finally, Fig.1c p o ides he e en s udy g aphs o la ge i ms. As be o e, he
CTA holds o mos speci ica ions. Only o in angible asse s in he se ice sec o
o acqui e s, simila ly o he eg essions wi h all i ms, we ind nega i e signi i-
can di e ences in p e- ends. The es ima ed e ec s o la ge i ms a e, on a e age,
s onge and main ain also longe han o small i ms. The only excep ion is sales,
whe e he sum o all h ee pe iods , + 1 and + 2 is he same as o small i ms.
All in all, we can summa ize ha we ind posi i e signi ican e ec s o “ a ge s” in
he manu ac u ing sec o on all ou economic a iables. The CTA seems o hold in
mos cases. The o e all esul s seems o be mainly d i en by la ge i ms.
We show he s ongly ela ed DiD eg ession esul s o he sho and long e m
wi h he cumula ed e ec s ac oss he h ee yea s a e in e na ionalisa ion in Appen-
dix Tables11, 12, 13, 14 and 15 (wi h he co esponding plo s in Figs.2 and 3 o
Table 5 E en s udy g aph—
es ima ed ea men e ec s
o “ a ge ed” i ms in he
manu ac u ing sec o
Signi ican ea men e ec s o “ a ge ed” i ms in e en s udy
g aphs
+ 1 + 2 Sum o ,
+ 1 and
+ 2
All i ms
P oduc i i y 5.1% 5.0% 3.2% 13.3%
Sales 7.7% 7.2% 6.5% 21.4%
In angible asse s 16.6% 17.4% 33.9%
Labou cos s 5.8% 5.9% 3.4% 15.1%
Small i ms
P oduc i i y 3.5% 4.3% 7.7%
Sales 6.2% 5.9% 5.9% 17.9%
In angible asse s 18.8% 18.8%
Labou cos s 4.9% 5.8% 10.7%
La ge i ms
P oduc i i y 6.2% 5.5% 4.5% 16.5%
Sales 6.1% 6.1% 5.6% 17.9%
In angible asse s 24.1% 24.1%
Labou cos s 6.0% 5.5% 4.6% 16.2%
929
In e na ionalisa ion asaboos o many i ms: e idence om…
p oduc i i y).41 The in e p e a ion o ou esul s o small and la ge i ms e e s o
he o al e ec o he acquisi ion, including he base e ec plus he e ec on ele an
in e ac ion e ms. As be o e, we examine he manu ac u ing and he se ices sec o s
sepa a ely. In ou explana ions, we concen a e on he ou pu ables as hey include
he cumula ed e ec s and as hey allow be e compa isons o he e ec s ac oss he
pe o mance a iables and ac oss he sec o s h ough he p o ision o exac igu es
o he e ec s.
4.1 Fi ms’ pos ‑acquisi ion pe o mance
4.1.1 Impac o owne ship change onini ially na ional i ms
Wi h espec o he p oduc i i y o i ms in he manu ac u ing sec o , he plo s o he
e en s udy al eady show posi i e e ec s, especially o la ge i ms. In acco dance,
ou DiD eg essions ind – a he 1% and 5% le el – signi ican ly posi i e sho - e m
and accumula ed long- e m e ec s d i en by la ge i ms ( o hese, he e ec s a e
signi ican a he 1% and 5% le el) (see AppendixTable11, columns 2 and 4). Fo
small i ms, his signi ican e ec is only obse ed in he sho e m. Thus, he new
o eign owne s a e capable o s imula ing la ge i ms, in pa icula , o inc ease hei
p oduc i i y. The e may be s onge o e laps be ween business ac i i ies o la ge
i ms wi h acqui e s ha may allow o e iciency gains h ough he implemen a ion
o be e echnology. We es ima e p oduc i i y e ec s o la ge i ms o 3.1% in he
sho un and 7.7% in he long un.42 Ou esul s a e quan i a i ely a he small com-
pa ed o he empi ical li e a u e. Guadalupe e al. (2012), Bi can (2019) and A nold
and Ja o czik (2009) documen la ge p oduc i i y e ec s o 11% o Spanish i ms,
13% o Tu kish plan s and 13.5% o i ms in Indonesia. This is easonable since
mos o he pape s concen a e on de eloping coun ies (Tu key, India, and Indone-
sia). Fu he mo e, hey add ess ea lie pe iods, whe e globalisa ion was mo e p o-
nounced (especially be ween 1985 and 2011). Fu he mo e, ou ou come con adic s
S iebale and Vencappa (2018). They p esume ha small manu ac u ing i ms a e
u he away om he echnological on ie han la ge ones and hus can lea n mo e
om an acqui e . They es ima e p oduc i i y e ec s o 14.8% o small Indian i ms.
We only es ima e e ec s o 2.7% in he sho un and no signi ican long- un e ec s
o small Ge man i ms. While ou classi ica ions o small is based on he sum o
angible and in angible asse s, hey e e o sales. Besides, hei sample o o eign
acquisi ions is a he small. In he case o i ms in he se ices sec o , we only see a
weakly posi i e sho - e m e ec and his ime only o all i ms (signi ican a he
10% le el in column5). In he se ices sec o , p oduc i i y g ow h seems o be less
41 To es ima e DiD in an unbiased way, we conduc ed he desc ibed empi ical s eps in Sec .3: i s , we
s a wi h p obi es ima ions o es ima e p opensi y sco es. The ollowing ma ching p o ides us wi h a
weigh ing scheme ha allows us o es ima e he e ec o a o eign acquisi ion on a ge i m pe o mance
using DiD in an app op ia e way. Ou DiD eg essions below ely on Eqs.(3) and (4) om Sec .3.
42 The e ec s sligh ly di e because he e en s udy app oach includes wo mo e ea men a iables
(s a s wi h k = −2 ins ead o k = 0). This leads o a d op in obse a ions (and possible sel -selec ion
e ec s).
930
R.F ey, S.Goldbach
a ainable. Thus, adjus men s wi hin a new a ilia e a e p esumably mo e bu den-
some in he se ices sec o , which no mally p ides i sel wi h close clien ela ions.
Fig. 2 E ec o in e na ionalisa ion on a ge s’ o al ac o p oduc i i y (le : manu ac u ing sec o ; igh :
se ices sec o )
Fig. 3 E ec o in e na ionalisa ion on acqui e s’ o al ac o p oduc i i y (le : manu ac u ing sec o ;
igh : se ices sec o )

931
In e na ionalisa ion asaboos o many i ms: e idence om…
In e na ionalisa ion p omo es sales, no only o small bu also o la ge i ms.
One main mo i e o going global is o bene i om economies o scale. Small and
la ge manu ac u ing i ms p o i om signi ican ly posi i e sho - e m e ec s on
sales (a he 1 and 5% signi icance le el), asAppendix Table12 shows. These posi-
i e e ec s a e, on a e age, s onge o small i ms. In he long e m, he e ec s
emain highly signi ican o small i ms, while signi icance anishes o some ex en
o la ge i ms. I is p obably he case ha in es ing companies choose i ms o c e-
a e new dis ibu ion channels o hei p e ious p oduc s – hus indica ing a ho -
izon al me ge (see Ma kusen (1984)). In he case o la ge-scale me ge s, ma ke
concen a ion e ec s may be in ended and he sales e ec on he a ious uni s o he
conglome a e may be ini ially unclea . Again, he economic e ec s a e smalle han
in he empi ical li e a u e. Guadalupe e al. (2012) documen s ong e ec s o 18%
o Spanish i ms. We ind 5.5% in he sho un and 15.6% in he long un. Fo he
se ices sec o , we see posi i e e ec s in he sho e m o small and la ge i ms (a
he 5% le el). He e, he impac is s onge o la ge i ms.
The pa e n la gely changes when we u n o he e ec s on in angible asse s
(see Appendix Table 13). He e, we see no e ec s wi hin he manu ac u ing sec-
o . Thus, ans e s o knowledge epo ed in he balance shee do no seem o be
ele an wi hin he ime span we add ess. Howe e , he pic u e is di e en o he
se ices sec o whe e we ind mixed esul s: la ge se ices i ms can be seen as
knowledge p o i ee s, expe iencing posi i e sho and long- e m e ec s (signi ican
a he 1% le el). Thus, i appea s ha in es o s s eng hen new a ilia es and ha
hey may es ablish new concep s (e.g. anchising). In con as , we ind signi ican ly
nega i e sho and long- e m e ec s o smalle i ms (a he 5% le el). In angible
asse s could be ans e ed o he new pa en company ab oad o in angibles asse s
could become useless and w i en o . The la e could be he case o adema ks,
o example.
Finally, AppendixTable14 shows posi i e sho and long- e m e ec s on labou
cos s applying o la ge and small i ms in he manu ac u ing sec o (a he 1 and 5%
signi icance le el in he sho e m and he 5 and 10% le el in he long e m). Ei he
employmen o wages inc eased – he la e pe haps going hand in hand wi h he
employmen o mo e highly quali ied s a . This esul would be in line wi h Hijzen
e al. (2013) who ind posi i e wage e ec s h ough employmen g ow h in highly
skilled jobs, wi h no e idence o g ea e job insecu i y. Egge e  al. (2020) also
p o ide easons o highe wages: he applica ion o new echnologies may equi e
wo ke aining, which is likely o be accompanied by highe wages. Fu he mo e, a
wage p emium may be paid o p o ec he echnological ad an age o he mul ina-
ional.43 Thus, he ea o job losses seems o be un ounded, and he opposi e may
ins ead be ue wi h espec o small manu ac u ing i ms ( hough i is also possible
ha , ac ually, wages inc eased). Con e sely, in he se ices sec o , we do no ind
any signi ican esul s. As highligh ed p e iously, he e a e many missing alues o
employmen in he da ase . Ne e heless, we conduc ed a obus ness check using
43 Acco ding o Koch and Smolka (2019), ou pu gains a e highes when i ms engage in bo h echnol-
ogy and skill upg ading a he same ime. Thus, new o eign owne s sha e p o i s wi h he highly skilled
wo ke s who a e al eady in he plan be o e he acquisi ion (see Bals ik and Halle (2010)).
932
R.F ey, S.Goldbach
employmen ins ead o labou cos s (seeAppendix Table12). We ind signi ican ly
posi i e esul s in he manu ac u ing sec o o la ge i ms only. Thus, he posi i e
labou cos e ec s could esul om highe employmen . We would in e p e hese
esul s wi h cau ion since we canno ule ou a po en ial selec ion bias.
Thus, ou s ong esul s wi h espec o he de elopmen o he pe o mance a i-
ables o ou Ge man manu ac u ing a ge i ms show ha o eign in es men s make
hei new Ge man a ilia es p o i o he mos pa . Howe e , in he case o he se -
ices sec o , he e idence is much weake and less signi ican . An explici empi i-
cal compa ison o he coe icien s o he wo sec o s shows only mino di e ences
(seeAppendix Table11).
4.1.2 Impac on i ms in es ing ab oad
AppendixTable15 p esen s he coe icien s o sho and long- e m e ec s o i s -
ime FDI on p oduc i i y o he manu ac u ing (uppe panel) and he se ices
(lowe panel) sec o s.44 An impac on he pe o mance o he pa en company would
no come as a su p ise as in e na ional in es men s a e likely o igge i m es uc-
u ing. On he o he hand, he acqui ing i m is, on a e age, la ge han he acqui ed
i m and i s i m s uc u e mo e complex. Thus, e ec s may be ela i ely smalle o
may ake mo e ime o un old.
Ou esul s in Table15 poin o he second s and o in e p e a ion, as mos e ec s
a e insigni ican . Wi h espec o p oduc i i y, in angible asse s and labou cos s, we
do no ind any signi ican e ec s. Thus, i may be a he bu densome o a leas
ime-consuming o an ini ially na ional en e p ise o es uc u e in such a way as o
p o i ei he h ough he exploi a ion o economies o scale by ocusing on p oduc-
ion p ocesses o h ough p ocess inno a ion in he newly es ablished c oss-bo de
alue o dis ibu ion chain. In e na ionalisa ion may ende he i m s uc u e mo e
complex.
We ind a leas some posi i e e ec s in he a e ma h o in e na ionalisa ion o
sales: small i ms inc ease hei sales in he sho e m in he manu ac u ing sec o
(a he 5% signi icance le el). In he se ices sec o , we gene ally see no posi i e
sales e ec s, indica ing ha he local ma ke con inues o be se ed by he newly
acqui ed a ilia e.
The s ock o in angible asse s is gene ally no a ec ed by expanding ab oad. The
small dynamic i ms pa icipa ing in in e na ionalisa ion p obably al eady ha e
some know-how in a pa icula a ea ha hey may in end o also exploi ab oad and,
a leas a he s a ing s age o he mul ina ional i m’s c ea ion, no u he e o s
44 The h ee di e en coe icien s o all, small and la ge i ms base on wo di e en eg essions.
In he ollowing, we explain how we compu e hem. The coe icien Changei (0.031) in Appen-
dixTable11 column (1) is equi alen o “All” (0.007) in AppendixTable15. This co esponds o β0 in
Eq.(3).Appendix Table11 column (2) dis inguishes hen be ween small and la ge i ms. The coe icien
Changei + Changei x Smalli (0.027) is he o al e ec o small “ a ge ” i ms. The co esponding e ec
is “Small” (0.003) in AppendixTable 15. This is equal o β0 + γ0 in Eq.(4). The coe icien Changei
(0.035) is he o al e ec o la ge “ a ge ” i ms. The co esponding e ec is “La ge” (0.012) in Appen-
dixTable15. This is equal o β0 in Eq.(4). All explana ions also hold o he long- e m e ec , whe e we
use ins ead o β0 and γ0 only, he sum o β0 + β1 + β2 and γ0 + γ1 + γ2.
933
In e na ionalisa ion asaboos o many i ms: e idence om…
leading o as bene i s, e.g. wi h espec o he pa en ’s echnological s ance, a e
obse ed.
In addi ion, we could no de ec any e ec on labou cos s o i ms s a ing o go
ab oad. Thus, a leas he e is no e idence o nega i e labou ou sou cing e ec s ha
may be de imen al o he domes ic labou o ce. Goldbach e al. (2019) ind wi h
espec o Ge man mul ina ionals ha new o eign a ilia es b ing wi h hem e en
mo e in es men a home and employmen inc eases as well. Howe e , Lich enbe g
and Siegel (1987) ind ha i ms sell plan s ha a e less p oduc i e. In addi ion,
Maksimo ic e al. (2011) ind ex ensi e es uc u ing in he a e ma h o acquisi-
ions – which is likely o inc ease he compe i i eness o a company. Fo Ge man
i ms engaging in FDI o he i s ime, we e en see signi ican ly posi i e employ-
men e ec s hough only o la ge companies in he manu ac u ing sec o in he
sho e m, asAppendix Table17 shows.
4.2 Cha ac e is ics o  i ms aking pa inin e na ionalisa ion
The p obi es ima ions, which ely on Eq.(2) om Sec .3, a e an econome ic p e-
equisi e o he DiD es ima ions in he p eceding Sec .4.1. We b ie ly look a he
ou come ables om hese es ima ions as hey e eal he speci ic cha ac e is ics o
i ms pa icipa ing in in e na ionalisa ion.
Fi s , he coe icien s o a ge s’ and acqui e s’ i m size (measu ed as log o o al
asse s) in he pe iod p eceding he acquisi ion a e signi ican ly posi i e a he 1%
le el (see Table4, columns (1)–(4) o bo h he manu ac u ing and he se ices sec-
o s). Fo eign in es o s may acqui e Ge man i ms o bene i om hei in as uc-
u e and epu a ion; con e sely, na ional i ms need economic esou ces o en e o -
eign ma ke s h ough FDI. Thus, la ge en e p ises may be gene ally be e equipped
o such an unde aking. Fu he mo e, wi h espec o po en ial long- e m e ec s,
Gugle e al. (2003) see la ge i m me ge s as a means o inc ease ma ke powe .
Tu ning o speci ic asse s, we ind ha in angible asse s a e o s ong ele ance
o a ge s in he manu ac u ing sec o and o acqui e i ms e en in bo h sec o s
– again wi h a signi ican ly posi i e coe icien a he 1% le el. The eason o his is
ha knowledge o b and-based ad an ages can also be applied ab oad. They would
allow o scale e ec s and ul ima ely esul in highe compe i i eness. Ai ken e al.
(1996) obse e ha mul ina ional i ms possess in angible p oduc i e asse s such as
echnological know-how, ma ke ing and managemen skills, expo con ac s, coo -
dina ed ela ionships wi h supplie s and cus ome s, and epu a ion. Addi ionally,
he in en ion behind a c oss-bo de i m buy-o may also be o ha ness addi ional
knowledge o boos p oduc i i y in he u u e, meaning ha he compe i i eness o
he conglome a e as a whole could inc ease (see also F ey and Hussinge (2011)).
By con as , we ind a signi ican ly nega i e coe icien o angible asse s in he case
o he se ices sec o . Thus, equipmen and buildings seem o be mo e o a bu den
han an ad an age o i ms wi h in e na ional ambi ions.
In a a he su p ising de ia ion om he li e a u e, in he case o he manu ac u -
ing sec o , i ms pa icipa ing in he in e na ionalisa ion p ocess a e less p oduc i e.
Also su p isingly, i ms in es ing ab oad o he i s ime also seem o ha e o he
934
R.F ey, S.Goldbach
compa a i e ad an ages. Acco dingly, acqui e s in his sec o ha e highe labou
cos s; his may indica e highe quali ica ion le els (see Egge e  al. (2020)) o a
la ge labou o ce.45
Empi ical e idence o he posi i e ele ance o sales is limi ed o a ge i ms
in he manu ac u ing sec o . Addi ionally, we ind signi ican ly nega i e esul s o
p o i abili y, de ined as e u n on equi y in he case o he se ices sec o .
Finally, in ou sample, he numbe o i ms in he se ices sec o ha in es
ab oad is conside ably lowe han in he manu ac u ing sec o (see Table2). On he
one hand, his unde lines he high impo ance o he Ge man manu ac u ing sec-
o ; on he o he hand, i aises ques ions abou he in e na ional s eng h o Ge man
i ms in he se ices sec o . This is e en mo e s iking gi en ha o eign acqui e s
in es ai ly equally in manu ac u ing and se ices i ms in Ge many (see Table1).
5 Robus ness
5.1 Fi ms’ in es men ab oad h oughM&A o ho izon al in es men
Ou empi ical se up i s examines acquisi ions o Ge man i ms by o eign com-
panies, ollowed by he o he in es men di ec ion ( i s - ime in es men ab oad by
Ge man i ms), which encompasses M&A and g een ield in es men s. This aises
he ques ion as o which ype o ou wa d in es men d i es ou baseline esul s. To
o e come his da a limi a ion, we p o ide se e al obus ness checks wi h di e en
app oxima ions o M&A ansac ions.
We s a wi h Ge man ou wa d ansac ions wi h o he indus ialised coun ies
(see AppendixTable18). The li e a u e a gues ha i ms a e mo e likely o con-
duc ho izon al FDI in high-income coun ies o se e o eign ma ke s, while i ms
engage in e ical FDI in low-income coun ies wi h lowe labou cos s. We iden i y
ho izon al FDI in coun ies wi hin he uppe qua ile o na ional GDP pe capi a
wi hin ou o al coun y sample. Desc ip i e s a is ics show ha , wi hin his qua ile,
68% o FDI is ho izon al, conduc ed by Ge man i ms going ab oad, p obably p e-
domina ely by means o M&A. We ind only insigni ican esul s, which con i ms
ou baseline indings.
Beside he dis inc ions g een ield s. M&A and ho izon al s. e ical FDI, he
empi ical li e a u e p oposes ano he way o dis inguish be ween ho izon al and e -
ical FDI. I he pa en company and i s o eign a ilia e a e ac i e in he same sec o ,
hen i is mo e likely ha he i m will engage in ho izon al FDI. Engaging in di -
e en sec o s is de ined as e ical FDI, since i ms engage in c oss-bo de supply
chains. We de ine ho izon al FDI as a Ge man pa en company and i s o eign a ili-
a e decla ing he same wo-digi sec o classi ica ion. Acco ding o his de ini ion,
we ind esul s di e ing om ou baseline es ima es, especially in he manu ac u ing
sec o (seeAppendix Table19). Manu ac u ing i ms engaging in FDI o he i s
ime also bene i om going in e na ional. The esul s o his in es men di ec ion
45 Addi ionally, Bals ik and Halle (2010) conclude ha o eign i ms pick la ge, high-wage and high-
p oduc i i y i ms, which is also la gely co obo a ed by he li e a u e.
941
In e na ionalisa ion asaboos o many i ms: e idence om…
Table 8 Sample composi ion o i e yea es ic ion (“ a ge s”)
Manu ac u ing Se ices
No es ic ions Res ic ions No es ic ions Res ic ions
Impo an 2-digi economic sec o s
Manu ac u e o machine y and equipmen n.e.c 203 (19.5%) 191 (19.4%)
Manu ac u e o compu e , elec onic and op ical p oduc s 119 (11.4%) 104 (10.6%)
Manu ac u e o chemicals and chemical p oduc s 102 (9.8%) 86 (8.7%)
Wholesale ade, excep o mo o ehicles and mo o cycles 328 (35.5%) 304 (33.8%)
Compu e p og amming, consul ancy and ela ed ac i i ies 96 (10.4%) 93 (10.3%)
Fi m size
Mean o al asse s (in hsd. €) 125,911 159,167 70,184 88,043
Mean o al sales (in hsd. €) 149,767 183,559 112,431 115,544
Mean employmen 459 531 304 373
Geog aphical egion
Asia 62 (6.0%) 57 (5.8%) 70 (7.6%) 47 (5.2%)
No h Ame ica 123 (11.8%) 112 (11.4%) 90 (9.7%) 78 (8.7%)
Wes e n Eu ope 812 (78.1%) 779 (79.0%) 737 (79.8%) 736 (81.8%)
Numbe o i ms 1,040 986 924 900

942
R.F ey, S.Goldbach
Table 9 Balancing p ope y es o “ a ge s” o o al ac o p oduc i i y
The s anda dised pe cen age bias is he pe cen age di e ence o he sample means in he ea ed and non- ea ed ( ull o ma ched) sub-samples as a pe cen age o he
squa e oo o he a e age o he sample a iances in he ea ed and non- ea ed g oups (see Rosenbaum and Rubin (1985))
Sample Mean es
T ea ed Con ol % Bias p alue
Manu ac u ing
Log o al asse si -1 Unma ched 10.16 8.62 90.9 0.000
Ma ched 10.20 10.18 1.4 0.688
Log in angible asse si -1 Unma ched 4.27 2.34 70.8 0.000
Ma ched 4.66 4.73 −2.5 0.551
Log TFPi -1 Unma ched 4.08 2.92 65.5 0.000
Ma ched 4.34 4.34 −0.5 0.875
Log labou cos si -1 Unma ched 8.88 7.47 84.6 0.000
Ma ched 9.08 9.09 −0.4 0.898
Log salesi -1 Unma ched 10.54 8.94 87.4 0.000
Ma ched 10.66 10.64 0.8 0.805
Re u n on equi yi -1 Unma ched 4.84 2.11 3.4 0.047
Ma ched 0.87 1.18 −0.4 0.791
Log ixed asse si -1 Unma ched 7.71 6.61 41.5 0.000
Ma ched 8.39 8.44 −1.6 0.591
TFP g ow hi -1 Unma ched 0.05 0.03 6.2 0.004
Ma ched 0.04 0.04 −1.2 0.730
943
In e na ionalisa ion asaboos o many i ms: e idence om…
Table 10 Balancing p ope y es o “ a ge s” o o al ac o p oduc i i y
The s anda dised pe cen age bias is he pe cen age di e ence o he sample means in he ea ed and non- ea ed ( ull o ma ched) sub-samples as a pe cen age o he
squa e oo o he a e age o he sample a iances in he ea ed and non- ea ed g oups ( o mulae om Rosenbaum and Rubin (1985))
Sample Mean es
T ea ed Con ol % Bias p alue
Se ices
Log o al asse si -1 Unma ched 10.16 8.62 90.9 0.000
Ma ched 9.75 9.64 6.3 0.170
Log in angible asse si -1 Unma ched 4.27 2.34 70.8 0.000
Ma ched 3.75 3.66 3.2 0.553
Log TFPi -1 Unma ched 4.08 2.92 65.5 0.000
Ma ched 3.54 3.45 4.7 0.284
Log labou cos si -1 Unma ched 8.88 7.47 84.6 0.000
Ma ched 8.47 8.36 6.1 0.209
Log salesi -1 Unma ched 10.54 8.94 87.4 0.000
Ma ched 10.30 10.23 3.6 0.422
Re u n on equi yi -1 Unma ched 4.84 2.11 3.4 0.047
Ma ched 1.67 2.75 −1.3 0.425
Log ixed asse si -1 Unma ched 7.71 6.61 41.5 0.000
Ma ched 6.92 6.90 0.7 0.889
TFP g ow hi -1 Unma ched 0.05 0.03 6.2 0.031
Ma ched 0.05 0.05 2.4 0.654
944
R.F ey, S.Goldbach
Table 11 E ec o in e na ionalisa ion on a ge s’ o al ac o p oduc i i y
OLS panel es ima ion. The dependen a iable is o al ac o p oduc i i y (in logs). The uni o obse a ion is i m-yea obse a ion. The sample pe iod is 1999 o 2018.
Fi m ixed e ec s and sec o - ime-speci ic ixed e ec s a e included bu no epo ed. Robus s anda d e o s (clus e ed by i m) a e in pa en heses. ***, ** and * deno e
signi icance a he 1%, 5% and 10% le el, espec i ely
Manu ac u ing Se ices
(1) (2) (3) (4) (5) (6) (7) (8)
Changei 0.031*** 0.035*** 0.030*** 0.034** 0.022* 0.029 −0.005 0.008
(0.008) (0.012) (0.011) (0.014) (0.013) (0.019) (0.018) (0.025)
Changei -1 0.030*** 0.025** 0.013 0.026
(0.010) (0.012) (0.018) (0.024)
Changei -2 0.014 0.016 0.012 0.032
(0.010) (0.011) (0.014) (0.020)
Changei x Smalli−0.008 −0.008 −0.014 −0.027
(0.017) (0.023) (0.025) (0.032)
Changei -1 × Smalli0.008 −0.028
(0.022) (0.031)
Changei -2 × Smalli−0.003 −0.039
(0.020) (0.026)
Changei + Changei x Smalli0.027** 0.015
p alue: 0.019 p alue: 0.373
2
∑
j
=0
Changei −
j
0.074** 0.075** 0.020 0.067
p alue: 0.011 p alue: 0.022 p alue: 0.664 p alue: 0.278
2
∑
j
=0
Changei −j+
2
∑
j
=0
Changei −j
x
Smalli
0.072 −0.027
p alue: 0.146 p alue: 0.654
Obse a ions 124,182 124,182 103,563 103,563 290,577 290,577 236,148 236,148
Adj. R2 0.943 0.943 0.953 0.953 0.953 0.953 0.966 0.966
945
In e na ionalisa ion asaboos o many i ms: e idence om…
Table 12 E ec o in e na ionalisa ion on a ge s’ sales
OLS panel es ima ion. The dependen a iable is o al sales (in logs). The uni o obse a ion is i m-yea obse a ion. The sample pe iod is 1999 o 2018. Fi m ixed
e ec s and sec o - ime-speci ic ixed e ec s a e included bu no epo ed. Robus s anda d e o s (clus e ed by i m) a e in pa en heses. ***, ** and * deno e signi icance
a he 1%, 5% and 10% le el, espec i ely
Manu ac u ing Se ices
(1) (2) (3) (4) (5) (6) (7) (8)
Changei 0.054*** 0.035** 0.053*** 0.029* 0.056*** 0.065** 0.023 0.023
(0.012) (0.014) (0.016) (0.017) (0.018) (0.027) (0.027) (0.037)
Changei -1 0.049*** 0.028* 0.037 0.025
(0.015) (0.017) (0.028) (0.041)
Changei -2 0.043*** 0.026 0.041** 0.050*
(0.015) (0.018) (0.019) (0.029)
Changei x Smalli0.037 0.051 −0.017 0.001
(0.024) (0.033) (0.035) (0.046)
Changei -1 × Smalli0.043 0.022
(0.032) (0.049)
Changei -2 × Smalli0.037 −0.017
(0.029) (0.040)
Changei + Changei x Smalli0.073*** 0.048**
p alue: 0.000 p alue: 0.041
∑2
j
=0Changei −
j
0.145*** 0.082* 0.101 0.098
p alue: 0.000 p alue: 0.069 p alue: 0.143 p alue: 0.321
∑2
j
=0Changei −j
+
∑2
j
=0Changei −
j
x Smalli
0.214*** 0.104
p alue: 0.004 p alue: 0.226
Obse a ions 126,443 126,443 105,366 105,366 301,720 301,720 244,756 244,756
Adj. R2 0.866 0.866 0.915 0.915 0.870 0.870 0.908 0.908
946
R.F ey, S.Goldbach
Table 13 E ec o in e na ionalisa ion on a ge s’ in angible asse s
OLS panel es ima ion. The dependen a iable is in angible asse s (in logs). The uni o obse a ion is i m-yea obse a ion. The sample pe iod is 1999 o 2018. Fi m
ixed e ec s and sec o - ime-speci ic ixed e ec s a e included bu no epo ed. Robus s anda d e o s (clus e ed by i m) a e in pa en heses. ***, ** and * deno e signi i-
cance a he 1%, 5% and 10% le el, espec i ely
Manu ac u ing Se ices
(1) (2) (3) (4) (5) (6) (7) (8)
Changei 0.005 −0.046 0.023 −0.009 0.042 0.244*** 0.030 0.313***
(0.042) (0.061) (0.055) (0.077) (0.049) (0.073) (0.070) (0.098)
Changei -1 0.038 0.057 0.049 0.291***
(0.055) (0.077) (0.065) (0.088)
Changei -2 −0.056 −0.046 0.056 0.251***
(0.053) (0.075) (0.060) (0.082)
Changei x Smalli0.099 0.061 −0.394*** −0.586***
(0.088) (0.112) (0.100) (0.136)
Changei -1 × Smalli−0.035 −0.506***
(0.110) (0.132)
Changei -2 × Smalli−0.019 −0.416***
(0.106) (0.127)
Changei + Changei x Smalli0.054 −0.150**
p alue: 0.377 p alue: 0.025
∑2
j
=0Changei −
j
0.005 0.002 0.136 0.856***
p alue: 0.973 p alue:0.992 p alue: 0.446 p alue: 0.001
∑2
j
=0Changei −j
+
∑2
j
=0Changei −
j
x Smalli
0.009 −0.654**
p alue: 0.966 p alue: 0.011
Obse a ions 126,310 126,310 105,270 105,270 301,266 301,266 244,350 244,350
Adj. R2 0.745 0.745 0.763 0.763 0.807 0.808 0.831 0.832

947
In e na ionalisa ion asaboos o many i ms: e idence om…
Table 14 E ec o in e na ionalisa ion on a ge s’ labou cos s
OLS panel es ima ion. The dependen a iable is labou cos s (in logs). The uni o obse a ion is i m-yea obse a ion. The sample pe iod is 1999 o 2018. Fi m ixed
e ec s and sec o - ime-speci ic ixed e ec s a e included bu no epo ed. Robus s anda d e o s (clus e ed by i m) a e in pa en heses. ***, ** and * deno e signi icance
a he 1%, 5% and 10% le el, espec i ely
Manu ac u ing Se ices
(1) (2) (3) (4) (5) (6) (7) (8)
Changei 0.035*** 0.037*** 0.033*** 0.032** 0.020 0.017 −0.019 −0.024
(0.009) (0.013) (0.012) (0.015) (0.017) (0.027) (0.022) (0.032)
Changei -1 0.035*** 0.025** −0.008 −0.027
(0.011) (0.013) (0.022) (0.035)
Changei -2 0.012 0.016 0.009 0.008
(0.012) (0.012) (0.016) (0.022)
Changei x Smalli−0.004 0.002 0.006 0.011
(0.020) (0.026) (0.034) (0.043)
Changei -1 × Smalli0.019 0.037
(0.023) (0.044)
Changei -2 × Smalli−0.008 0.004
(0.023) (0.033)
Changei + Changei x Smalli0.033** 0.023
p alue: 0.016 p alue: 0.251
∑2
j
=0Changei −
j
0.079*** 0.073** −0.018 −0.043
p alue: 0.009 p alue: 0.037 p alue: 0.744 p alue: 0.577
∑2
j
=0Changei −j
+
∑2
j
=0Changei −
j
x Smalli
0.086* 0.008
p alue: 0.099 p alue: 0.912
Obse a ions 126,698 126,698 105,622 105,622 294,650 294,650 239,250 239,250
Adj. R2 0.936 0.936 0.947 0.947 0.929 0.929 0.944 0.944
948
R.F ey, S.Goldbach
Table 15 E ec o in e na ionalisa ion on acqui e s’ p oduc i i y, sales, in angible asse s and labou cos s
Coe icien s o in e es wi h clus e ed s anda d e o s (by i ms) in pa en heses
Sho - e m e ec s Long- e m e ec
Log TFP Log sales Log in angible asse s Log labou cos s Log TFP Log sales Log in angible asse s Log labou cos s
Manu ac u ing
All 0.007 0.026** −0.058 0.006 −0.009 0.020 −0.136 −0.014
(0.008) (0.010) (0.046) (0.008) (0.029) (0.035) (0.159) (0.027)
Small 0.003 0.033** −0.073 −0.003 −0.001 0.038 −0.209 −0.023
(0.011) (0.016) (0.055) (0.011) (0.039) (0.050) (0.186) (0.041)
La ge 0.012 0.018 −0.042 0.017 −0.018 0.003 −0.064 −0.005
(0.011) (0.014) (0.074) (0.012) (0.040) (0.048) (0.258) (0.042)
Obse a ions 67,275 68,290 68,244 68,861 56,716 57,520 57,457 57,978
Adj. R20.934 0.876 0.667 0.931 0.949 0.925 0.693 0.948
Se ices
All 0.013 0.003 −0.133* 0.020 0.067 0.020 −0.146 0.079
(0.014) (0.025) (0.072) (0.017) (0.061) (0.100) (0.245) (0.067)
Small −0.001 0.002 −0.103 −0.007 0.062 0.030 −0.319 0.022
(0.016) (0.028) (0.109) (0.018) (0.077) (0.101) (0.392) (0.084)
La ge 0.026 0.004 −0.163* 0.046 0.072 0.010 −0.001 0.128
(0.025) (0.041) (0.094) (0.032) (0.095) (0.163) (0.304) (0.104)
Obse a ions 228,207 236,931 241,579 234,086 187,024 193,856 197,378 191,550
Adj. R20.967 0.897 0.824 0.957 0.976 0.922 0.846 0.967
949
In e na ionalisa ion asaboos o many i ms: e idence om…
Table 16 Di e ences be ween manu ac u ing and se ices
Coe icien s o in e es wi h clus e ed s anda d e o s (by i ms) in pa en heses
Sho - e m e ec s Long- e m e ec s
Log TFP Log sales Log in angible asse s Log labou cos s Log TFP Log sales Log in angible asse s Log labou cos s
Ta ge s
All 0.005 0.002 −0.060 0.008 0.053 0.070 −0.175 0.073
(0.016) (0.022) (0.066) (0.019) (0.056) (0.080) (0.235) (0.063)
Small 0.008 0.060 0.407*** 0.015 0.112 0.187* 0.279 0.138
(0.034) (0.047) (0.136) (0.045) (0.080) (0.107) (0.325) (0.093)
La ge −0.006 −0.033 −0.323*** −0.004 −0.064 −0.073 −0.954*** −0.034
(0.027) (0.037) (0.097) (0.038) (0.083) (0.129) (0.348) (0.096)
Obse a ions 410,273 421,507 420,963 415,903 336,748 345,557 345,184 341,155
Adj. R20.952 0.878 0.777 0.932 0.963 0.916 0.799 0.946
Acqui e s
All −0.008 −0.037 0.095 −0.022 −0.079 −0.216* −0.004 −0.108
(0.017) (0.035) (0.084) (0.018) (0.067) (0.130) (0.288) (0.075)
Small −0.039 0.004 −0.142 −0.021 −0.143 −0.225 −0.105 −0.167*
(0.037) (0.065) (0.170) (0.045) (0.089) (0.159) (0.388) (0.096)
La ge 0.012 −0.038 0.177 −0.013 −0.015 −0.212 0.104 −0.054
(0.029) (0.058) (0.127) (0.038) (0.099) (0.186) (0.414) (0.115)
Obse a ions 301,468 312,836 307,183 306,547 248,539 257,330 253,024 252,439
Adj. R20.957 0.857 0.730 0.934 0.968 0.913 0.755 0.949
950
R.F ey, S.Goldbach
Table 17 Employmen as dependen a iable
Coe icien s o in e es wi h clus e ed s anda d e o s (by i ms) in pa en heses
Ta ge s Acqui e s
Manu ac u ing Se ices Manu ac u ing Se ices
Sho - e m Long- e m Sho - e m Long- e m Sho - e m Long- e m Sho - e m Long- e m
All 0.031*** 0.072* 0.015 0.013 0.017 −0.020 0.010 0.009
(0.011) (0.038) (0.027) (0.089) (0.020) (0.058) (0.035) (0.132)
Small 0.024 0.027 −0.021 −0.035 −0.025 −0.089 −0.008 −0.131
(0.018) (0.062) (0.033) (0.137) (0.024) (0.056) (0.043) (0.164)
La ge 0.037*** 0.108** 0.048 0.048 0.065** 0.053 0.034 0.174
(0.013) (0.045) (0.036) (0.092) (0.030) (0.101) (0.052) (0.200)
Obse a ions 76,901 66,066 133,892 113,652 40,202 34,665 118,854 100,643
Adj. R20.964 0.969 0.972 0.975 0.856 0.870 0.928 0.933
957
In e na ionalisa ion asaboos o many i ms: e idence om…
Table 24 Ge man pa en companies wi h i s - ime FDI (coa sened exac ma ching)
Coe icien s o in e es wi h clus e ed s anda d e o s (by i ms) in pa en heses
Sho - e m e ec s Long- e m e ec s
Log TFP Log sales Log in angible asse s Log labou cos s Log TFP Log sales Log in angible
asse s
Log labou cos s
Manu ac u ing
All 0.002 0.015 0.027 0.003 −0.015 0.026 0.011 −0.014
(0.008) (0.014) (0.050) (0.011) (0.025) (0.037) (0.179) (0.033)
Small 0.004 0.032 0.015 0.013 −0.020 0.030 −0.048 −0.018
(0.011) (0.020) (0.071) (0.013) (0.041) (0.054) (0.283) (0.048)
La ge 0.001 0.003 0.036 −0.005 −0.012 0.023 0.048 −0.012
(0.010) (0.020) (0.070) (0.017) (0.032) (0.049) (0.233) (0.044)
Obse a ions 12,462 13,179 13,182 12,365 10,472 11,154 11,094 10,461
Adj. R20.962 0.903 0.786 0.954 0.970 0.943 0.782 0.965
Se ices
All −0.007 −0.002 −0.135 0.015 −0.025 0.081 −0.127 0.005
(0.017) (0.035) (0.086) (0.019) (0.049) (0.113) (0.306) (0.061)
Small −0.008 0.036 −0.051 −0.002 0.011 0.090 0.055 0.030
(0.016) (0.038) (0.115) (0.021) (0.061) (0.142) (0.412) (0.084)
La ge −0.006 −0.038 −0.213* 0.030 −0.056 0.073 −0.294 −0.018
(0.022) (0.063) (0.118) (0.030) (0.071) (0.176) (0.422) (0.087)
Obse a ions 13,855 12,846 15,191 11,721 11,245 10,062 12,291 9,403
Adj. R20.984 0.884 0.881 0.976 0.986 0.889 0.880 0.978

958
R.F ey, S.Goldbach
Table 25 Fo eign akeo e s o Ge man companies (“ a ge ” is ne e in con ol g oup)
Coe icien s o in e es wi h clus e ed s anda d e o s (by i ms) in pa en heses
Sho - e m e ec s Long- e m e ec s
Log TFP Log sales Log in angible asse s Log labou cos s Log TFP Log sales Log in angible asse s Log labou cos s
Manu ac u ing
All 0.031** 0.060*** 0.068 0.028*** 0.076* 0.158*** 0.168 0.079*
(0.011) (0.016) (0.057) (0.011) (0.041) (0.061) (0.199) (0.041)
Small 0.031* 0.078*** 0.098 0.022 0.094 0.265** 0.175 0.092
(0.018) (0.026) (0.087) (0.020) (0.078) (0.110) (0.291) (0.077)
La ge 0.030* 0.044** 0.041 0.034*** 0.061 0.074 0.163 0.068
(0.014) (0.019) (0.077) (0.013) (0.043) (0.062) (0.271) (0.045)
Obse a ions 118,553 120,506 120,294 120,960 98,740 100,294 100,131 100,708
Adj. R20.943 0.868 0.742 0.941 0.953 0.916 0.762 0.951
Se ices
All 0.026 0.083*** 0.093 0.021 −0.011 0.070 0.349 −0.016
(0.017) (0.029) (0.067) (0.024) (0.063) (0.102) (0.250) (0.075)
Small 0.028 0.072** −0.155* 0.040 −0.018 0.136 −0.420 0.061
(0.023) (0.036) (0.093) (0.033) (0.084) (0.123) (0.353) (0.113)
La ge 0.021 0.094** 0.359*** −0.001 −0.003 0.005 1.091*** −0.103
(0.025) (0.042) (0.099) (0.036) (0.083) (0.150) (0.345) (0.103)
Obse a ions 285,346 296,271 295,662 289,244 231,714 240,146 239,608 234,670
Adj. R20.953 0.860 0.802 0.926 0.966 0.902 0.832 0.940
959
In e na ionalisa ion asaboos o many i ms: e idence om…
Table 26 Ge man pa en companies wi h i s - ime FDI (“acqui e ” is ne e in con ol g oup)
Coe icien s o in e es wi h clus e ed s anda d e o s (by i ms) in pa en heses
Sho - e m e ec s Long- e m e ec s
Log TFP Log sales Log in angible asse s Log labou cos s Log TFP Log sales Log in angible asse s Log labou cos s
Manu ac u ing
All −0.005 0.016 −0.048 −0.007 −0.047 −0.022 −0.052 −0.060
(0.012) (0.013) (0.063) (0.013) (0.041) (0.046) (0.205) (0.042)
Small −0.008 0.019 0.022 −0.020 −0.036 −0.033 0.103 −0.078
(0.018) (0.019) (0.075) (0.019) (0.059) (0.074) (0.238) (0.061)
La ge −0.003 0.015 −0.105 0.005 −0.057 −0.013 −0.170 −0.045
(0.016) (0.018) (0.100) (0.017) (0.056) (0.057) (0.321) (0.057)
Obse a ions 63,328 64,259 64,198 64,806 53,264 53,997 53,925 54,435
Adj. R20.932 0.887 0.678 0.930 0.948 0.932 0.702 0.949
Se ices
All 0.040 0.033 −0.092 0.055 0.041 0.131 −0.015 0.037
(0.032) (0.060) (0.107) (0.034) (0.115) (0.197) (0.297) (0.120)
Small −0.029 −0.072 −0.086 −0.060 −0.041 −0.031 −0.448 −0.200
(0.043) (0.086) (0.210) (0.045) (0.184) (0.259) (0.431) (0.175)
La ge 0.078* 0.087 −0.095 0.116** 0.080 0.207 0.179 0.147
(0.044) (0.079) (0.122) (0.046) (0.145) (0.265) (0.385) (0.151)
Obse a ions 225,798 234,443 239,052 231,661 184,936 191,702 195,191 189,448
Adj. R20.965 0.878 0.820 0.949 0.975 0.906 0.848 0.962
960
R.F ey, S.Goldbach
Table 27 S agge ed DiD wi h Callaway and San ’Anna (2021) es ima o
Coe icien s o in e es wi h clus e ed s anda d e o s (by i ms) in pa en heses
Manu ac u ing Se ices
Log TFP Log sales Log in angible asse s Log labou cos s Log TFP Log sales Log in angible asse s Log labou cos s
Ta ge s
All −0.003 0.011 −0.080 −0.008 0.022 0.111*** 0.073 0.023
(0.009) (0.013) (0.057) (0.011) (0.014) (0.038) (0.061) (0.016)
Small 0.005 0.033* −0.071 −0.003 0.023 0.158** 0.012 0.018
(0.013) (0.018) (0.089) (0.014) (0.017) (0.077) (0.087) (0.020)
La ge −0.010 −0.011 −0.075 −0.012 0.022 0.072*** 0.131 0.028
(0.012) (0.017) (0.070) (0.015) (0.021) (0.025) (0.084) (0.023)
Acqui e s
All 0.035*** 0.028** 0.208*** 0.033*** 0.048** 0.062 0.142 0.055**
(0.011) (0.014) (0.069) (0.012) (0.019) (0.044) (0.098) (0.026)
Small 0.053*** 0.050*** 0.225*** 0.047** 0.079** 0.135 0.308** 0.071**
(0.014) (0.015) (0.078) (0.015) (0.035) (0.084) (0.138) (0.028)
La ge 0.008 −0.006 0.157 0.010 0.018 −0.003 −0.001 0.035
(0.015) (0.022) (0.107) (0.017) (0.029) (0.048) (0.140) (0.041)
961
In e na ionalisa ion asaboos o many i ms: e idence om…
Funding Open Access unding enabled and o ganized by P ojek DEAL.
Decla a ions
Con lic o in e es The au ho s decla e ha hey ha e no known compe ing inancial in e es s o pe sonal
ela ionships ha could ha e appea ed o in luence he wo k epo ed in his pape .
Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long
as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e
Commons licence, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his
a icle a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line
o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended
use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mis-
sion di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/
licenses/by/4.0/.
Re e ences
Acke be g, D. A., Ca es, K., & F aze , G. (2015). Iden i ica ion p ope ies o ecen p oduc ion unc-
ion es ima o s. Econome ica, 83(6), 2411–2451.
Ai ken, B., Ha ison, A., & Lipsey, R. E. (1996). Wages and o eign owne ship: a compa a i e s udy
o Mexico, Venezuela, and he Uni ed S a es. Jou nal o In e na ional Economics, 40, 345–371.
A nold, J. M., & Ja o cik, B. S. (2009). Gi ed kids o pushy pa en s? Fo eign di ec in es men and
plan p oduc i i y in Indonesia. Jou nal o In e na ional Economics, 79, 42–53.
Ash a , D., He ze , D., & Nunnenkamp, P. (2016). The e ec s o g een ield FDI and c oss-bo de
M&As on o al ac o p oduc i i y. Wo ld Economy, 39(11), 1728–1755.
Bake , A. C., La cke , D. F., & Wang, C. (2022). How much should we us s agge ed di e ence-in-
di e ences es ima es? Jou nal o Financial Economics, 144(2), 370–395.
Bals ik, R., & Halle , S. (2010). Picking “Lemons” o Picking “Che ies”? Domes ic and o eign
acquisi ions in No wegian manu ac u ing. Scandina ian Jou nal o Economics, 112(2), 361–387.
Bals ik, R., & Halle , S. (2020). Wo ke -plan ma ching and owne ship change. Scandina ian Jou nal
o Economics, 122(4), 1286–1314.
Becke , S. O., & Muendle , M.-A. (2010). Ma gins o mul ina ional labo subs i u ion. Ame ican Eco-
nomic Re iew, 100(5), 1999–2030.
Be and, O., Hakkala, K. N., No bäck, P.-J., & Pe sson, L. (2012). Should coun ies block o eign
akeo e s o R&D champions and p omo e g een ield en y? Canadian Jou nal o Economics,
45(3), 1083–1124.
Be and, O., & Zi ouna, H. (2008). Domes ic e sus c oss-bo de acquisi ions: Which impac on he
a ge i ms’ pe o mance? Applied Economics, 40(17), 2221–2238.
Bialek, S., & Weichen iede , A. J. (2021). Do s ingen en i onmen al policies de e FDI? M&A e -
sus g een ield. En i onmen al & Resou ce Economics, 80, 603–636.
Bi can, C. (2019). Owne ship s uc u e and p oduc i i y o mul ina ionals. Jou nal o In e na ional
Economics, 116, 125–143.
Caliendo, M., Kopeinig, S. (2005). Some p ac ical guidance o he implemen a ion o p opensi y
sco e ma ching. IZA Discussion Pape Se ies No 1588, May 2005.
Callaway, B., & San Anna, P. H. C. (2021). Di e ence-in-di e ences wi h mul iple ime pe iods.
Jou nal o Econome ics, 225(2), 200–230.
Cassiman, B., Colombo, M. G., Ga one, P., & Veugele s, R. (2005). The impac o M&A on he R&D
p ocess: an empi ical analysis o he ole o echnological- and ma ke - ela edness. Resea ch
Policy, 34(2), 195–220.
Ca es, R. E. (1974). Mul ina ional i ms, compe i ion, and p oduc i i y in hos -coun y ma ke s. Eco-
nomica, 41(162), 176–193.
962
R.F ey, S.Goldbach
Chen, W. (2011). The e ec o in es o o igin on i m pe o mance: domes ic and o eign di ec
in es men in he Uni ed S a es. Jou nal o In e na ional Economics, 83(2), 219–228.
Con educa, F., Kazako a, E. (2018). Se ing ab oad: Expo , M&A, and g een ield in es men . CRC
TR224 Discussion Pape Se ies No. 008, Ma ch 2018.
Conyon, M. J., Gi ma, S., Thompson, S., & W igh , P. W. (2002). The p oduc i i y and wage e ec s o
o eign acquisi ion in he Uni ed Kingdom. The Jou nal o Indus ial Economics, 50(1), 85–102.
C ass, D., & Pe e s, B. (2014). In angible asse s and i m-le el p oduc i i y. SSRN Elec onic Jou nal.
h ps:// doi. o g/ 10. 2139/ ss n. 25623 02
C ouze , N., Ebe ly, J. (2019). Unde s anding weak capi al in es men : The ole o ma ke concen a-
ion and in angibles. NBER Wo king Pape No 25869.
Da ies, R., Ma kusen, J. R. (2020). The s uc u e o mul ina ional i ms’ in e na ional ac i i ies.
NBER Wo king Pape No. 26827, Ma ch 2020.
Da ies, R. B., Desbo des, R., & Ray, A. (2018). G een ield e sus me ge and acquisi ion FDI: Same
wine, di e en bo les? Canadian Jou nal o Economics/ e ue Canadienne D’économique,
51(4), 1151–1190. h ps:// doi. o g/ 10. 1111/ caje. 12353
Da is, S., Hal iwange , J., Handley, K., Ja min, R., Le ne , J., & Mi anda, J. (2014). P i a e equi y,
jobs, and p oduc i i y. Ame ican Economic Re iew, 104(12), 3956–3990.
Dehejia, R., & Wahba, S. (2002). P opensi y sco e-ma ching me hods o nonexpe imen al causal
s udies. The Re iew o Economics and S a is ics, 84(1), 151–161.
Desai Hines, M. A., Foley, C. F., & James, R. (2009). Domes ic e ec s o he o eign ac i i ies o US.
Ame ican Economic Jou nal: Economic Policy, 1(1), 181–203.
Egge , H., Jahn, E., & Ko ni zky, S. (2020). Reassessing he o eign owne ship wage p emium in Ge -
many. The Wo ld Economy, 43(2), 302–325.
Egge , H., & K eickemeie , U. (2013). Why o eign owne ship may be good o you. In e na ional
Economic Re iew, 54(2), 693–716.
Fa ell, J., & Shapi o, C. (1990). Ho izon al me ge s: an equilib ium analysis. Ame ican Economic
Re iew, 80(1), 107–126.
F ey, R., & Hussinge , K. (2011). Eu opean ma ke in eg a ion h ough echnology-d i en M&As.
Applied Economics, 43(17), 2143–2153.
Fues , C., Hugge , F., Sul an, S., & Xing, J. (2021). Wha d i es Chinese o e seas M&A in es men ?
E idence om mic o da a. Re iew o In e na ional Economics, Online:
Gandhi, A., Na a o, S., & Ri e s, D. (2020). On he iden i ica ion o g oss ou pu p oduc ion unc-
ions. Jou nal o Poli ical Economy, 128(8), 2973–3016.
Ge s enbe ge , J. (2018). Mo e Chinese M&A deals in he Ge man SME sec o —bu he sha e
emains negligible. K W Resea ch Pape 229.
Gi ma, S., & Gö g, H. (2007). E alua ing he o eign owne ship wage p emium using a di e ence-in-
di e ences ma ching app oach. Jou nal o In e na ional Economics, 72(1), 97–112.
Goldbach, S., Nagengas , A., S einmülle , E., & Wamse , G. (2019). The e ec o in es ing ab oad on
in es men a home: on he ole o echnology, ax sa ings, and in e nal capi al ma ke s. Jou nal
o In e na ional Economics, 116, 58–73.
Guadalupe, M., Kuzmina, O., & Thomas, C. (2012). Inno a ion and o eign owne ship. Ame ican
Economic Re iew, 102(7), 3594–3627.
Gugle , K., Muelle , D., Yu oglu, B., & Zulehne , C. (2003). The e ec s o me ge s: an in e na ional
compa ison. In e na ional Jou nal o Indus ial O ganiza ion, 21(5), 625–653.
Ha ms, P., & Méon, P.-G. (2018). Good and useless FDI: he g ow h e ec s o g een ield in es men
and me ge s and acquisi ions. Re iew o In e na ional Econonomics, 26, 37–59.
Helpman, E. (1984). A simple heo y o in e na ional ade wi h mul ina ional co po a ions. Jou nal
o Poli ical Economy, 92(3), 451–471.
Helpman, E., Meli z, M., & Yeaple, S. (2004). Expo e sus FDI wi h he e ogeneous i ms. Ame ican
Economic Re iew, 91(4), 300–316.
Henna , J.-F., & Pa k, Y.-R. (1993). G een ield e sus acquisi ion: he s a egy o Japanese in es o s
in he Uni ed S a es. Managemen Science, 39(9), 1054–1070.
Hijzen, A., Ma ins, P. S., Schank, T., & Upwa d, R. (2013). Fo eign-owned i ms a ound he wo ld: a
compa a i e analysis o wages and employmen a he mic o-le el. Eu opean Economic Re iew,
60, 170–188.
Iacus, S. M., King, G., & Po o, G. (2011). Mul i a ia e ma ching me hods ha a e mono onic imbal-
ance bounding. Jou nal o he Ame ican S a is ical Associa ion, 106(493), 345–361.

963
In e na ionalisa ion asaboos o many i ms: e idence om…
Iacus, S. M., King, G., & Po o, G. (2012). Causal in e ence wi hou balance checking: coa sened
exac ma ching. Poli ical Analysis, 20(1), 1–24.
Ja e, A. B., T aj enbe g, M., & Hende son, R. (1993). Geog aphic localiza ion o knowledge spillo-
e s as e idenced by pa en ci a ions. The Qua e ly Jou nal o Economics, 108(3), 577–598.
Ja o cik, B., & Poelhekke, S. (2017). Fo me o eign a ilia es: Cas ou and ou pe o med? Jou nal o
he Eu opean Economic Associa ion, 15(3), 501–539.
Koch, M., & Smolka, M. (2019). Fo eign owne ship and skill-biased echnological change. Jou nal o
In e na ional Economics, 118, 84–104.
Le insohn, J., & Pe in, A. (2003). Es ima ing p oduc ion unc ions using inpu s o con ol o unob-
se ables. Re iew o Economic S udies, 70(2), 317–341.
Lich enbe g, F. R., Siegel, D., Jo genson, D., & Mans ield, E. (1987). P oduc i i y and changes in
owne ship o manu ac u ing plan s. B ookings Pape s on Economic Ac i i y, 1987(3), 643–683.
Maksimo ic, V., Phillips, G., & P abhala, N. R. (2011). Pos -me ge es uc u ing and he bounda ies
o he i m. Jou nal o Financial Economics, 102(2), 317–343.
Ma kusen, J. (1984). Mul ina ionals, mul i-plan economies, and he gains om ade. Jou nal o
In e na ional Economics, 16(3), 205–226.
Nocke, V., & Yeaple, S. (2007). C oss-bo de me ge s and acquisi ions s. g een ield o eign di ec
in es men . The ole o i m he e ogenei y. Jou nal o In e na ional Economics, 72(2), 336–365.
h ps:// doi. o g/ 10. 1016/j. jin e co. 2006. 09. 003
Regula ion (EU) (2019). Es ablishing a amewo k o he sc eening o o eign di ec in es men s in o he
Union. O icial Jou nal o he Eu opean Union L, 79, 1–14.
Ro h, J., San ’Anna, P. H. C., Bilinski, A., & Poe, J. (2023). Wha ’s ending in di e ence-in-di e ences?
A syn hesis o he ecen econome ics li e a u e. Jou nal o Econome ics, 235(2), 2218–2244.
S iebale, J., & Reize, F. (2011). The impac o FDI h ough me ge s and acquisi ions on inno a ion in
a ge i ms. In e na ional Jou nal o Indus ial O ganiza ion, 29(2), 155–167.
S iebale, J., & Vencappa, D. (2018). Acquisi ions, ma kups, e iciency, and p oduc quali y: e idence
om India. Jou nal o In e na ional Economics, 112, 70–87.
Szücs, F. (2014). M&A and R&D: asymme ic e ec s on acqui e s and a ge s? Resea ch Policy, 43(7),
1264–1273.
Teece, D. J. (1977). Technology ans e by mul ina ional i ms: he esou ce cos o ans e ing echno-
logical know-how. Economic Jou nal, 87(346), 242–261.
Van Bieseb oeck, J. (2008). The sensi i i y o p oduc i i y es ima es. Jou nal o Business & Economic
S a is ics, 26, 311–328.
Wang, J., & Wang, X. (2015). Bene i s o o eign owne ship: e idence om o eign di ec in es men in
China. Jou nal o In e na ional Economics, 97(2), 325–338.
Woold idge, J. M. (2009). On es ima ing i m-le el p oduc ion unc ions using p oxy a iables o con ol
o unobse ables. Economics Le e s, 104(3), 112–114.
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps
and ins i u ional a ilia ions.