F ode mann, Co inna; W ohlich, Ka ha ina; Zucco, Aline
A icle — Accep ed Manusc ip (Pos p in )
Pa en al Lea e Policy and Long- un Ea nings o Mo he s
Labou Economics
P o ided in Coope a ion wi h:
Ge man Ins i u e o Economic Resea ch (DIW Be lin)
Sugges ed Ci a ion: F ode mann, Co inna; W ohlich, Ka ha ina; Zucco, Aline (2023) : Pa en al Lea e
Policy and Long- un Ea nings o Mo he s, Labou Economics, ISSN 1879-1034, Else ie , Ams e dam,
Vol. 80, pp. 1-13,
h ps://doi.o g/10.1016/j.labeco.2022.102296
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/308497
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-nc-nd/4.0/
Pa en al Lea e Policy and
Long- un Ea nings o Mo he s∗
Co inna F ode mann1, Ka ha ina W ohlich†2, and Aline Zucco3
1Ins i u e o Employmen Resea ch, Nu embe g, Ge many
2Ge man Ins i u e o Economic Resea ch (DIW Be lin) and Uni e si y o Po sdam, Ge many
3Hans Boeckle Founda ion
No embe 2022
Abs ac : Paid pa en al lea e schemes ha e been shown o inc ease women’s employmen a es bu
o dec ease hei wages in case o ex ended lea e du a ion. In iew o hese po en ial ade-o s, many
coun ies a e discussing he op imal design o pa en al lea e policies. We analyze he impac o a majo
pa en al lea e e o m on mo he s’ long- e m ea nings. The 2007 Ge man pa en al lea e e o m eplaced a
means- es ed bene i wi h a mo e gene ous ea nings- ela ed bene i ha is g an ed o a sho e pe iod o
ime. Addi ionally, a ”daddy quo a” o wo mon hs was in oduced. To iden i y he causal e ec o his
policy mix on long- un ea nings o mo he s, we use a di e ence-in-di e ences app oach ha compa es
labo ma ke ou comes o mo he s who ga e bi h jus be o e and igh a e he e o m and ne s ou
seasonal e ec s by including he yea be o e. Using adminis a i e social secu i y da a, we con i m
p e ious indings and show ha he a e age du a ion o employmen in e up ions inc eased o mo he s
wi h high p e-bi h ea nings. Ne e heless, we ind a posi i e long- un e ec on ea nings o mo he s in
his g oup. This e ec canno be explained by changes in he selec ion o wo king mo he s, wo king hou s
o changes in employe s abili y. Desc ip i e e idence sugges s ha he s onge in ol emen o a he s,
incen i ized by he ”daddy mon hs”, could ha e acili a ed mo he s’ e-en y in o he labo ma ke and
he eby inc eased ea nings. Fo mo he s wi h low p e-bi h ea nings, howe e , we do no ind bene icial
long- un e ec s o his pa en al lea e e o m.
JEL: H31, J13, J22, J24, J31
Keywo ds: Pa en al lea e, wages, labo supply
∗We hank he Hans Boeckle Founda ion o unding o he esea ch p ojec N . 2017-327-3. Addi ionally, we
would like o hank he colleagues om he Resea ch Da a Cen e o he Ins i u e o Employmen Resea ch in
Nu embe g and o he Fede al S a is ical O ice in Be lin.
†Co esponding au ho . [email p o ec ed]
This is he pos p in o an a icle published in: Labou Economics 80 (2023), 102296, 13 S.,
a ailable online a : h ps://doi.o g/10.1016/j.labeco.2022.102296
© <2023>. This manusc ip e sion is made a ailable unde he CC-BY-NC-ND 4.0
license h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/
1 In oduc ion
Paid pa en al lea e as i is common in many indus ialized coun ies has been shown o be
associa ed wi h highe women’s employmen a es on he one hand, bu lowe ela i e wages a
ex ended du a ions o paid lea e on he o he hand (e.g. Ruhm, 1998). A syn hesis o many
empi ical s udies om a ious coun ies and ins i u ional se ings shows ha he e seems o
be a non-mono onic ela ionship be ween he leng h o he lea e and mo he s’ labo ma ke
ou comes (e.g. Oli e i and Pe ongolo, 2017). Besides p omo ing wo k- amily li e balance,
amily policy o en de ines addi ional goals such as child well-being and inancial secu i y o
amilies ha migh con lic wi h he goal o s eng hening he labo ma ke a achmen and,
he eby, he economic independence o mo he s. Agains he backg ound o a ious po en ial
ade-o s, many coun ies a e cu en ly discussing he op imal design o pa en al lea e policies.
Ge many implemen ed a pa en al lea e e o m in 2007 ha changed he old pa en al lea e
bene i scheme in h ee impo an ways. Fi s , i eplaced a means- es ed bene i a ge ed a
lowe -income amilies by an ea nings ela ed ans e ha is paid o all amilies. Second, while
he du a ion o job-p o ec ion (36 mon hs) has no been changed, he maximum du a ion o paid
lea e was cu om 24 o 12 mon hs. Depending on household income and indi idual p e-bi h
ea nings, his changed wo k incen i es o mo he s in he i s and second yea a e gi ing bi h.
Incen i es o ake a lea e o one yea ha e s ongly inc eased o mo he s wi h high p e-bi h
ea nings. Fo mo he s wi h low p e-bi h ea nings and low household income, incen i es o ake
up employmen in he second yea a e gi ing bi h ha e inc eased. Thi d, he new pa en al
lea e scheme in oduced a ”daddy quo a” o wo mon hs. This means ha wo ou o 14 mon hs
a e ea ma ked indi idually o each pa en . I one (non-single) pa en does no ake pa en al
lea e, he maximum du a ion o paid pa en al lea e o he amily is wel e mon hs.
Se e al empi ical e alua ion s udies ha e shown ha his e o m had he expec ed sho - un
labo supply e ec s o mo he s: Labo supply o mo he s wi h high p e-bi h ea nings dec eased
in he i s yea a e gi ing bi h, whe eas labo supply o mo he s wi h low p e-bi h ea nings
inc eased in he second yea a e gi ing bi h (see, e.g. Be gemann and Riphahn, 2011,2015;
Geye , Haan, and W ohlich, 2015; Klu e and Schmi z, 2018; Klu e and Tamm, 2013; Wel eke
and W ohlich, 2019). Much less, howe e , is known abou he medium- and long- un e ec s
o his pa en al lea e e o m. One s udy by Klu e and Schmi z (2018) has shown ha he
pa en al lea e e o m posi i ely a ec ed job cha ac e is ics such as du a ion o con ac in he
medium un, in pa icula o high-income mo he s, in he medium- un. E alua ion s udies ha
analyzed p e ious e o ms o ma e ni y lea e in (Wes ) Ge many om he 1970s, 1980s and
1990s ha e shown ha ex ensions o paid lea e and he job-p o ec ed lea e ha e inc eased he
employmen in e up ions o mo he s (Sch¨onbe g and Luds eck, 2014). These longe employmen
1
in e up ions had – wi h one excep ion – no nega i e e ec s on mo he s’ ea nings in he medium
e m (i.e. up o six yea s a e gi ing bi h).1
Ou pape con ibu es o he li e a u e in wo ways. Fi s , we b oaden he ocus om
he sho and medium- e m pe spec i e o long- e m ou comes, i.e. up o nine yea s a e
gi ing bi h. Second, we in es iga e he e ec s o he 2007 pa en al lea e e o m on mo he s’
ea nings - an ou come ha has no been analyzed in his con ex so a . To his end, we
use adminis a i e da a om he In eg a ed Employmen Biog aphies (IEB) o he Ins i u e o
Employmen Resea ch (IAB) ha con ain in o ma ion on he o al popula ion o indi iduals
in Ge many who ha e an employmen con ac subjec o social secu i y con ibu ions. Fo
he iden i ica ion o he causal e ec o he pa en al lea e e o m o 2007 on mo he s’ ea nings
we ollow p e ious li e a u e on sho - e m e ec s o pa en al lea e in Ge many and exploi
he quasi-expe imen ha was p o ided by he unan icipa ed in oduc ion o he pa en al lea e
bene i in Janua y 2007. In pa icula , we compa e mo he s whose i s child was bo n in he
las qua e o 2006 (con ol g oup) o mo he s whose i s child was bo n in he i s qua e
o 2007 ( ea men g oup). In o de o ake seasonal di e ences in o accoun , we add i s - ime
mo he s who ga e bi h o a child in he las qua e o 2005 and he i s qua e o 2006 and
employ a di e ence-in-di e ences app oach.
Ou esul s con i m p e ious indings and show ha he pa en al lea e e o m has inc eased
employmen in e up ions o mo he s wi h high p e-bi h ea nings by almos h ee mon hs on
a e age. Howe e , hese longe employmen in e up ions did no lead o lowe ea nings in he
long un. On he con a y, we ind ha wo o nine yea s a e gi ing bi h, hese mo he s
ha e highe ea nings. This esul is ob ained bo h, in a sample o all mo he s who ha e been
employed p io o gi ing bi h as well as in a sample o mo he s e u ning o wo k a e gi ing
bi h. Fo mo he s wi h low p e-bi h ea nings, on he o he hand, he esul s a e di e en : we
do no ind posi i e e ec s on ea nings in he medium and longe un (i.e. 5 o 9 yea s a e
gi ing bi h).
Based on u he empi ical analyses, we a e able o ule ou ha he posi i e ea nings
e ec s o mo he s wi h high p e-bi h ea nings a e caused by a change in wo king hou s, socio-
demog aphic cha ac e is ics o wo king mo he s, changes in e ili y pa e ns o changes in em-
ploye s abili y.
One po en ial mechanism causing he inc ease in mo he s’ ea nings in he i s couple o
yea s a e gi ing bi h could be he inc eased sha e o a he s ge ing in ol ed in childca e due
o he e o m. Desc ip i e e idence shows ha a he s whose pa ne has high ea nings had he
s onges eac ion o he in oduc ion o he “daddy quo a” and s a ed aking pa en al lea e
1The e o m o 1986 ex ended he ma e ni y bene i pe iod beyond he job p o ec ion pe iod and is he one
ha caused he s onges educ ion in ma e nal employmen . I is also he only e o m ha has lowe ed mo he s’
ea nings 6 yea s a e childbi h. (Sch¨onbe g and Luds eck, 2014) conclude ha job p o ec ion seems o be a e y
impo an de e minan o mo he sˆa pos bi h ca ee s.
2
a much highe a es han he a e age. P e ious s udies ha e shown ha pa e ni y lea e can
inc ease a he s’ in ol emen in childca e and has he po en ial o educe gende specializa ion
in ma ke wo k and unpaid ca e wo k. Thus, we hypo hesize ha he inc eased in ol emen o
a he s in pa en al lea e could ha e acili a ed mo he s’ e-en y in o he labo ma ke , inc eased
hei p oduc i i y and he eby also inc eased hei ea nings.
F om a policy pe spec i e, ou esul s show ha g an ing a mo e gene ous bene i in he
i s yea in o de o p o ide a inancial sa egua d o amilies wi h young child en oge he
wi h in oducing a ‘daddy quo a’ wi hin he pa en al lea e scheme has inc eased he du a ion o
employmen in e up ions o ce ain g oups o mo he s wi hou ha ming hei long- e m ca ee
pe spec i es. In con as , mo he s wi h high p e-bi h ea nings e en expe ience posi i e e ec s
on hei ea nings. Thus, he suspec ed ade-o be ween p o iding a sa egua d o amilies
wi h a new-bo n child and s eng hening mo he s’ labo ma ke a achmen and hei long- e m
economic independence does no seem o be empi ically ele an in he con ex o he Ge man
pa en al lea e e o m o 2007. In ac , he e is sugges i e e idence ha his ade-o could ha e
been mi iga ed by simul aneously incen i izing he use o pa en al lea e by a he s, he eby
acili a ing he e-en y in o he labo ma ke o mo he s a e hei amily- ela ed employmen
in e up ions.
Howe e , ou esul s also e eal ha only mo he s wi h medium o high incomes bene i ed
om his pa en al lea e e o m in e ms o highe long- e m ea nings. Low-income mo he s
po en ially los income due o he cu o he maximum du a ion pe iod, and hey did no gain
highe ea nings in he medium o long un.
The pape is o ganized as ollows. In he nex sec ion we desc ibe he ins i u ional se ing
and summa ize he ela ed li e a u e. Sec ion 3p esen s he empi ical app oach, while sec ion
4p o ides in o ma ion on he da a. We p esen he esul s o ou empi ical analysis in sec ion
5and sec ion 6concludes.
2 Ins i u ional Backg ound and P e ious Li e a u e
In Ge many, pa en al lea e legisla ion is a he gene ous wi h espec o job p o ec ion and
mone a y bene i s. Fi s , he e is ma e ni y lea e, which assu es employed women a lea e o
six weeks be o e and eigh weeks a e gi ing childbi h and inancial bene i s ha eplace hei
o al ne p io - o-bi h ea nings. A e his, each pa en can ake pa en al lea e om his o he
job and is g an ed employmen p o ec ion o a maximum o h ee yea s. Howe e , no all o
his maximum pa en al lea e pe iod is o has been paid: Un il he end o 2006, amilies wi h a
newbo n child could d aw a cash bene i amoun ing o 300 eu o pe mon h o a maximum pe iod
o 24 mon hs (E ziehungsgeld). This bene i was means- es ed a he household le el and income
h esholds we e se o a ge he median o household income o amilies wi h young child en.
3
Abo e his income h eshold, amilies did no ge any inancial bene i a e he ma e ni y lea e
pe iod expi ed.2
On Janua y 1s 2007, Ge many implemen ed a majo pa en al lea e e o m ha had h ee
goals. Fi s , i was mean o inc ease inancial s abili y o amilies wi h young child en and
p o iding a inancial sa egua d du ing he i s yea o a child’s li e. Second, an explici aim
s a ed in he law was o inc ease economic independence o bo h pa en s, in pa icula mo he s’,
by sho ening employmen in e up ions o mo he s. Finally, gende equali y goals we e also
explici ly s a ed in he law: Fa he s should be encou aged o ake a mo e ac i e ole in child
ca e by in oducing inancial incen i es such as a a he s’ quo a in he pa en al lea e bene i s
scheme.
The El e ngeld ha was in oduced o all child en bo n on o a e Janua y 1s 2007 eplaced
he E ziehungsgeld. In con as o his p e ious bene i , he new El e ngeld is no means- es ed
and mo e gene ous o mos amilies. I eplaces 67 pe cen o p io - o-bi h ne ea nings o he
pa en on lea e, up o a maximum o 1,800 eu o pe mon h. The minimum amoun o El e ngeld
awa ded is 300 eu o pe mon h, which is equi alen o he mon hly bene i paid unde he
p e ious E ziehungsgeld. Howe e , i is paid o a sho e pe iod o ime (12 mon hs i only one
pa en akes lea e o 14 mon hs i bo h pa en s ake lea e).
Figu e 1summa izes he changes in inancial incen i es o wo exempli ying mo he s who
ea n 1,000 eu o and 3,000 eu o pe mon h, espec i ely. Depending on p io - o-bi h ea nings and
household income, he e o m changed inancial incen i es o wo k in a di e en way du ing he
i s wo yea s a e he child is bo n. Fo mo he s wi h high p io - o-bi h ea nings, incen i es
o s ay a home in he i s yea a e he ma e ni y lea e pe iod inc eased s ongly. Fo mo he s
wi h no o low p io - o-bi h ea nings and below-median ne household income, incen i es o go
back o wo k in he i s yea dec eased, while hey inc eased in he second yea .
The pa en al lea e bene i e o m o 2007 has been e alua ed in se e al empi ical s udies.
Fo example, Klu e and Tamm (2013) exploi he quasi-expe imen al se ing o he e o m
and ind ha he p obabili y ha mo he s e u n o wo k du ing he i s yea a e gi ing
bi h o a child has declined, in pa icula o high-income mo he s. This inding has been
con i med in la e s udies, e.g. by Be gemann and Riphahn (2011,2015), Geye e al. (2015),
and Wel eke and W ohlich (2019). Klu e and Schmi z (2018) analyze mo he s’ employmen
esponses no only in he i s and second yea a e gi ing bi h bu also in he hi d o i h
yea . Based on da a om he Ge man Mic ocensus, hey ind a la ge and signi ican inc ease in
he employmen a e o mo he s wi h h ee o i e yea old child en. Mo eo e , hey ind ha he
e o m inc eased employe con inui y, i.e. a highe sha e o mo he s e u ned o hei p e-bi h
2Ac ually, he e we e wo di e en income h esholds a ound he median income. I household income exceeded
he highe h eshold (22,086 Eu o ne household income pe yea in 2006), he bene i was wi hd awn comple ely.
Below he lowe h eshold (16,500 Eu o ne household income pe yea in 2006), i was g an ed ully. In be ween
hese wo h esholds, he bene i was diminished by 5.2 pe cen o he income exceeding he lowe h eshold.
4
employe . Mo eo e , Wel eke and W ohlich (2019) show ha he e o m changed social no ms
ega ding he leng h o pa en al lea e ia social in e ac ion e ec s among cowo ke s.3
Figu e 1: Changes in inancial incen i es due o he 2007 pa en al lea e e o m
Sou ce: Wel eke and W ohlich (2019)
Resea ch analyzing se e al pa en al and ma e ni y lea e e o ms om ea lie pe iods (Lud-
s eck, 2014) has shown ha ex ending pa en al lea e in he 1970s, 1980s and 1990s in (Wes )
Ge many p olonged mo he s’ employmen in e up ions, howe e , did no a ec mo he s’ ea n-
ings in he medium un, i.e. up o 6 yea s a e gi ing bi h. Only one e o m ha s ongly
ex ended he pe iod o paid lea e ( om 6 o 22 mon hs) has been shown o dec ease medium-
e m ea nings o a small ex en . Simila ly, Lali e, Schlosse , S einhaue , and Zweim¨ulle (2014)
ha e shown o Aus ia ha e o ms ha ha e inc eased he maximum du a ion o paid lea e
in combina ion wi h job-p o ec ion ha e p olonged mo he s’ employmen in e up ions qui e
s ongly, howe e did no ha m mo he s’ ea nings in he medium un. Simila esul s ha e
also been ound o he pa en al lea e scheme in oduced in Cali o nia (e.g Baum and Ruhm,
2016), Canada (Bake and Milligan, 2008) and Aus alia (B oadway, Kalb, McVica , and Ma -
in, Fo hcoming). As summa ized by Rossin-Sla e (2018), he gene al conclusion om he
li e a u e is ha lea e en i lemen s up o one yea can imp o e job con inui y o women and in-
c ease hei labo ma ke a achmen , howe e , longe lea es can nega i ely a ec hei ea nings,
employmen and ca ee ad ancemen .
3The e a e many u he s udies analyzing he e ec o he 2007 pa en al lea e e o m wi h espec o o he
ou comes. Fo example, Cygan-Rehm (2016) and Rau e (2019) analyze i s e ec s on e ili y, Huebene , Kuehnle,
and Spiess (2019) he e ec s on child ou comes, Cygan-Rehm, Kuehnle, and Riphahn (2018) look a pa en s’
li ing a angemen s, and Tamm (2019) e alua es he e ec s on a he ’s childca e in ol emen .
5
Agains he backg ound o p e ious empi ical indings and he way how he 2007 pa en al
lea e e o m in Ge many changed incen i es o wo k du ing he i s and second yea a e
childbi h, we expec he ollowing e ec s on ea nings: Gi en ha he e o m has ambiguous
e ec s on he du a ion o employmen in e up ion o mo he s wi h low p e-bi h ea nings, la e
labo ma ke ou comes o his g oup could be ei he ze o, posi i e o nega i e. On he o he
hand, he expec ed longe employmen in e up ions o mo he s wi h high p e-bi h ea nings
esul ing om he pa en al lea e e o m, could po en ially lead o nega i e e ec s on long- e m
labo ma ke ou comes o his g oup. Howe e , since i has been shown by p e ious esea ch
(Klu e and Schmi z, 2018) ha his e o m has yielded some posi i e labo ma ke e ec s o
high-income mo he s in he medium un, such as highe employe s abili y and a la ge sha e o
pe manen wo k con ac s, he e migh also be posi i e e ec s on ea nings in he medium and
long un.
Due o hese ambiguous mechanisms, he sign and he magni ude o he long- e m e ec s o
he 2007 pa en al lea e e o m on mo he s’ ea nings emain an empi ical ques ion ha will be
analyzed in he emainde o his pape .
3 Empi ical App oach
To iden i y he causal e ec s o he e o m on mo he s’ long- un labo ma ke ou comes, we
exploi he in oduc ion o he new pa en al lea e bene i scheme in Janua y 2007 as a na u al
expe imen . Due o he iming o his e o m, pa en s o child en bo n in he i s qua e o
2007 could no know ha hey would be eligible o he new bene i s a he ime o concep ion
o hei child (see, e.g. Klu e and Tamm, 2013). Compa ing mo he s wi h child en bo n in he
i s qua e o 2007 ( ea men g oup) o mo he s wi h child en bo n in he las qua e o 2006
(con ol g oup) hus iden i ies he in en ion- o- ea e ec (ITT) o he e o m.
Mo he s wi h child en bo n in he las qua e o a yea , howe e , migh di e in hei labo
ma ke ou comes om mo he s wi h child en bo n in he i s qua e o a yea .4To con ol
o hese po en ial seasonal e ec s, we add obse a ions om he las qua e o 2005 and he
i s qua e o 2006 and employ a di e ence-in-di e ences es ima ion s a egy. Mo eo e , his
app oach allows con olling o po en ial seasonali y in he labo demand o o seasonal bonus
paymen s. In pa icula , we es ima e he e ec o he pa en al lea e e o m using he ollowing
equa ion:
Yi =α+βFi s Qua e i0+γRe o mi0+δFi s Qua e i0·Re o mi0+ωXi +ϵi (1)
whe e Yi deno es he log o he daily wage o mo he i in yea . The dummy a iable Fi s Qua -
e akes on he alue 1 i he mo he has gi en bi h o he i s child in he i s qua e o he
4P e ious s udies o he U.S. and he Czech Republic ha e shown ha he season o bi h is co ela ed wi h
socio-demog aphic ac o s o he mo he and he child’s la e ou comes (Bobak and Gjonca, 2001; Buckles and
Hunge man, 2013; Cla ke, O e ice, and Quin ana-Domeque, 2019).
6
yea 2006 o 2007 and 0 i he bi h has aken place in he las qua e o 2005 o 2006. The
dummy a iable Re o m akes on he alue 1 i he mo he ga e bi h o a child in he mon hs
a ound he implemen a ion o he e o m, i.e. in he las qua e o 2006 o in he i s qua e o
2007, and 0 i he bi h has aken place in he yea be o e. Unde ou iden i ying assump ions,
he coe icien δo he in e ac ion e m o hese wo dummy a iables is he causal e ec o he
pa en al lea e e o m. Figu e 2shows he de ini ion o he ea men and con ol g oups in ou
se ing g aphically.
Figu e 2: Depic ion o ea men and con ol g oups
2005 2006 2007
10 11 12 1 2 3 10 11 12 1 2 3 mon hs
Fi s Qua e Las Qua e
Con ols T ea ed
One yea be o e he
e o m (P e-Re o m)
Yea o he e-
o m (Re o m)
Sou ce: Own depic ion
Fo he iden i ica ion o he causal e ec o he pa en al lea e e o m on mo he s ea nings
we make use o he ac ha he e o m was announced in June 2006 and came in o e ec in
Janua y 2007 (Klu e and Tamm, 2013). This iming implies ha pa en s whose child was bo n
in he i s h ee mon hs o 2007 could no know ha hey would be eligible o he new bene i s
a he ime o concep ion o hei child. To exclude po en ial selec ion in o he pe iod a e he
cu -o da e by pos poning bi h5we exclude all mo he s who ha e gi en bi h o a child 14 days
be o e and a e Janua y 1.
4 Da a
Fo he empi ical analysis we mainly use indi idual in o ma ion gene a ed om labo admin-
is a ion o he Ge man Fede al Employmen Agency and om social secu i y da a p ocessing
(IEB) based on he in eg a ed no i ica ion p ocedu e o heal h, pension, and unemploymen
insu ances. The IEB is p o ided by he IAB and con ains he o al popula ion o indi iduals
in Ge many who ha e ei he an employmen con ac subjec o social secu i y con ibu ions,
5Neuga and Ohlsson (2013) and Tamm (2013) show ha some mo he s, in pa icula employed mo he s,
pos poned hei bi h a e Janua y 1s , 2007. The iming o bi h a ound cu -o poin s has also been ound o
o he e o ms (Dicke -Conlin and Chand a, 1999; Gans and Leigh, 2009)
7
Table 3: E ec s o he pa en al lea e e o m on daily ea nings o mo he s wi h high and low
p e-bi h ea nings
Low-income mo he s High-income mo he s
Daily
ea nings
Con ol Di e ence Di e ence Con ol Di e ence Di e ence
mean1T-C T-C mean1T-C T-C
wi h con ols2wi h con ols2
y -2 18.21 −0.467 −0.310 83.89 −0.863 −1.159*
(0.410) (0.342) (0.793) (0.476)
y -1 20.75 −0.011 0.145 91.89 1.455* 0.234
(0.347) (0.293) (0.694) (0.411)
y +1 4.18 −0.626** −0.559*** 20.64 −8.916*** −1.478*
(0.223) (0.210) (0.804) (0.631)
y +2 8.95 1.737*** 1.368*** 46.80 1.437 1.707*
(0.334) (1.009) (0.307) (0.685)
y +3 12.16 1.296*** 0.914* 48.06 2.224* 1.883**
(0.392) (0.359) (1.020) (0.709)
y +4 15.46 1.035* 0.850* 50.55 3.554*** 1.909**
(0.432) (0.394) (1.033) (0.691)
y +5 17.39 0.524 0.195 53.79 2.024* 1.315
(0.460) (0.423) (1.033) (0.693)
y +6 18.66 0.449 0.095 57.70 1.007 1.579*
(0.485) (0.447) (1.020) (0.696)
y +7 19.94 1.070* 0.791 61.79 −0.019 1.389*
(0.026) (0.470) (1.008) (0.678)
y +8 21.01 0.970 0.770 64.30 0.466 1.518*
(0.026) (0.490) (1.016) (0.689)
y +9 24.14 1.009 0.704 69.63 1.013 2.174**
(0.026) (0.533) (1.075) (0.727)
Sou ce: IEB 1976-2016; own calcula ions. All speci ica ions con ol o seasonal ends (p e- e o m
dummy), 1 Con ol mean e e s o he a e age mean o mo he s who ga e bi h in he las qua e
o 2006, as daily wage; Con ols 2: P e-bi h wage, age a bi h, educa ion, expe ience ( & p ), el.
du a ion o unemploymen , size o es ablishmen , wo king ime be o e bi h, change o es ablishmen ,
eas Ge many, ci izenship, no. o child en, egion, enu e and change o employe a e bi h. The numbe
o obse a ions is 38,631 o low-income mo he s and 42,563 o high-income mo he s. Signi icance le els:
* p<0.05, ** p<0.01, *** p<0.001.
14
Figu e 4: E ec s o he pa en al lea e e o m on daily ea nings o mo he s wi h high and low
p e-bi h ea nings (in Eu o)
-2 -1 1 2 3 4 5 6 7 8 9
−3
−2
−1
0
1
2
3
Yea s be o e/a e bi h
Low-income mo he s
High-income mo he s
Sou ce: IEB 1976-2016; own calcula ions. The g aph plo s he causal
e ec o he e o m wi h he co esponding 95% con idence in e als.
a e gi ing bi h. In sum, he esul s o medium-income mo he s esemble hose o low-income
mo he s: posi i e ea nings e ec s in he medium- un, bu no e ec s on ea nings in he long un.
To sum up, al hough he pa en al lea e e o m p olonged employmen in e up ions o a
la ge g oup o mo he s, his did no ha m hei ea nings in he yea s a e gi ing bi h. In
con as o p edic ions o human capi al heo y, we ind ha mo he s ge highe ea nings o
se e al yea s a e he bi h o hei child. Fo mo he s wi h high p e-bi h ea nings, hese
posi i e e ec s e en show up in he long- un, i.e. up o nine yea s a e gi ing bi h. In he nex
sub-sec ions, we analyze po en ial mechanisms ha could explain he posi i e long- un e ec s
on ea nings o mo he s wi h high p e-bi h income, in pa icula po en ial changes in wo king
hou s, socio-demog aphic cha ac e is ics o wo king mo he s o job cha ac e is ics.
5.3 E ec s on Wo king Hou s
One explana ion o he posi i e e ec s o he pa en al lea e e o m on he daily ea nings o
mo he s wi h high p e-bi h ea nings could be ha hey ha e longe wo king hou s a e hey
e u n o he labo ma ke . In o de o analyze whe he mo he s eac o he pa en al lea e
e o m wi h espec o hei wo king hou s, we es ima e equa ion (1) wi h he weekly wo king
hou s as he dependen a iable based on he Ge man Mic ocensus.
Reg ession esul s based on his da a se show ha he e is no s a is ically signi ican e ec
o he pa en al lea e e o m on weekly wo king hou s o mo he s in he yea s a e gi ing bi h
(Table 5). This is ue o bo h, mo he s wi h a high le el o o mal educa ion, and o hose
wi h a low le el o o mal educa ion.
wo k in he i s yea a e childbi h, al hough hey we e eligible o he bene i , ep esen a highly labo ma ke
a ached g oup o women.
15
Table 4: E ec o he pa en al lea e e o m on daily ea nings o mo he s wi h medium p io -
o-bi h ea nings
Medium-income mo he s
Con ol Di e ence Di e ence
mean1 T-C T-C
wi h con ols2
Du a ion o employmen b eak
(mon h)
25.00 1.038*
(0.412)
y −256.34 −0.481 −0.489
(0.380) (0.261)
y −160.75 −0.168 −0.268
(0.326) (0.216)
y +1 7.73 −2.720*** −0.686**
(0.266) (0.220)
y +2 19.04 2.129*** 1.641***
(0.371) (0.284)
y +3 23.54 1.062** 0.607*
(0.402) (0.306)
y +4 27.47 1.154** 0.847**
(0.416) (0.309)
y +5 28.40 0.517 0.332
(0.423) (0.322)
y +6 30.28 0.269 0.264
(0.426) (0.330)
y +7 32.51 0.325 0.161
(0.427) (0.329)
y +8 34.25 0.410 0.344
(0.437) (0.338)
y +9 38.00 0.639 0.156
(0.468) (0.361)
Sou ce: IEB 1976-2016; own calcula ions. All speci ica ions con ol o seasonal
ends (p e- e o m dummy), 1 Con ol mean e e s o he a e age mean o mo he s
who ga e bi h in he las qua e o 2006, as daily ea nings; Con ols 2: P e-bi h
wage, age a bi h, educa ion, expe ience ( & p ), el. du a ion o unemploymen ,
size o es ablishmen , wo king ime be o e bi h, change o es ablishmen , eas
Ge many, ci izenship, no. o child en, egion, enu e and change o employe a e
bi h. Numbe o obse a ions: 81,140. Signi icance le els: * p<0.05, ** p<0.01,
*** p<0.001. Robus s anda d e o s in pa en heses.
16
Based on his esul , we conclude ha he pa en al lea e e o m had no e ec on he wo king
hou s o mo he s a e e-en e ing he labo ma ke . Thus, he posi i e e ec on daily ea nings
o high-income mo he s does no s em om an ex ension o hei wo king hou s bu can a he
be in e p e ed as a posi i e e ec on hei hou ly wage.
Table 5: E ec o he pa en al lea e e o m on wo king hou s
Low-educa ed mo he s High-educa ed mo he s
Weekly
wo king hou s
Con ol Di e ence Di e ence Con ol Di e ence Di e ence
mean1T-C3T-C mean 1T-C3T-C
wi h con ols34 wi h con ols34
2008 24.78 6.340 5.525 29.40 1.444 1.408
P(β(T ea men ) =0)20.108 0.142 0.199 0.249
2009 23.32 1.586 0.175 30.20 −0.265 −0.592
P(β(T ea men ) =0)20.660 0.958 0.897 0.940
2010 23.29 1.924 3.176 29.18 1.481 1.371
P(β(T ea men ) =0)20.590 0.345 0.564 0.440
2011 26.28 1.630 2.075 29.28 0.430 0.721
P(β(T ea men ) =0)20.659 0.440 0.870 0.651
2012 23.66 2.805 2.948 28.64 −0.650 −1.269
P(β(T ea men ) =0)20.666 0.362 0.847 0.460
2013 22.29 −3.967 −2.630 28.16 0.778 0.365
P(β(T ea men ) =0)20.224 0.676 0.416 0.676
2014 20.22 1.422 2.656 27.94 0.533 0.523
P(β(T ea men ) =0)20.630 0.351 0.842 0.818
2015 18.88 2.891 4.314 27.45 −2.306 −1.891
P(β(T ea men ) =0)20.361 0.154 0.123 0.104
2016 24.53 −3.138 −3.630 28.56 0.150 −0.128
P(β(T ea men ) =0)20.225 0.300 0.475 0.580
Sou ce: RDC o he Fede al S a is ical O ice and S a is ical O ices o he L˜
A¤nde , Mic ocensus 2008-2016; own
calcula ions. All speci ica ions con ol o seasonal ends (p e- e o m dummy); T ea men dummy equals 1 i he i s
child is bo n in he i s hal o 2007; 1: The con ol mean equals he a e age wo king hou s o women who ga e bi h
o hei i s child in he las hal o 2006. 2: F-Tes o join signi icance: Fo low-educa ed P( β(T ea men )) and o
high-educa ed P( β(T ea men ) +β(T ea men * High-educa ed)) 3The size o he coe icien equals he sum o he
ea men e ec and he in e ac ion o he ea men and he highes educa ional g oup. 4Con ols: Age (single and
quad a ic), numbe o child en, egion (Eas s. Wes and U ban s. Ru al), na ionali y, ma i al s a us; The numbe
o obse a ions a y be ween 905 (2008) and 1,920 (2016). Signi icance le els: * p<0.05, ** p<0.01, *** p<0.001.
Robus s anda d e o s in pa en heses.
5.4 Changes in Socio-Demog aphic Cha ac e is ics
Since we do no obse e any s a is ically signi ican changes in weekly wo king hou s o mo h-
e s ha could explain he posi i e e ec s on daily ea nings, we analyze whe he he selec ion
o wo king mo he s in e ms o obse able socio-demog aphic cha ac e is ics has changed due
o he pa en al lea e e o m. To his end, we un se e al es ima ions o equa ion (1) wi h
17
socio-demog aphic cha ac e is ics such as educa ion le el, age and ma i al s a us as dependen
a iable. As summa ized in Table 6, he e is no s a is ically signi ican change in he composi-
ion o wo king mo he s wi h espec o hese cha ac e is ics esul ing om he pa en al lea e
e o m in any o he yea s 2008 o 2016, wi h only one mino excep ion. In he yea 2013, we
ind ha wo king mo he s in he ea men g oup ha e a highe p obabili y o be ma ied han
in he con ol g oups. We do no ind his esul , howe e , in any o he yea , and, he e o e,
a gue ha his inding is negligible.
Table 6: E ec s o he pa en al lea e e o m on he socio-demog aphic cha ac e is ics o wo king
mo he s
Yea
Va iable 2008 2009 2010 2011 2012 2013 2014 2015 2016
High educa ion
Con ol mean 0.63 0.64 0.67 0.70 0.67 0.67 0.69 0.74 0.75
Di e ence T-C 0.00 −0.04 −0.07 −0.02 0.05 0.07 0.05 0.02 −0.02
S anda d E o (0.05) (0.05) (0.05) (0.04) (0.04) (0.04) (0.04) (0.04) (0.03)
Age
Con ol mean 30.30 31.53 32.39 33.25 34.32 35.12 35.84 37.32 38.67
Di e ence T-C −0.17 −0.57 −0.45 0.60 0.90 0.72 0.66 −0.04 0.48
S anda d E o (0.61) (0.59) (0.57) (0.54) (0.54) (0.55) (0.51) (0.50) (0.45)
Ma ied
Con ol mean 0.72 0.74 0.75 0.75 0.75 0.73 0.74 0.76 0.76
Di e ence T-C 0.00 −0.07 −0.08 0.02 0.03 0.07 0.02 0.02 0.03
S anda d E o (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.03)
High-educa ed
Age
Con ol mean 33.55 34.52 35.27 36.04 37.10 37.96 38.10 39.22 40.46
Di e ence T-C −0.44 −0.29 −0.34 0.22 −0.13 0.16 0.45 −0.02 0.54
S anda d E o (0.53) (0.54) (0.52) (0.49) (0.51) (0.51) (0.51) (0.49) (0.45)
Ma ied
Con ol mean 0.80 0.80 0.84 0.81 0.82 0.80 0.79 0.79 0.79
Di e ence T-C −0.05 −0.01 −0.05 0.03 0.01 0.10* 0.02 0.02 0.05
S anda d E o (0.05) (0.05) (0.05) (0.04) (0.04) (0.04) (0.04) (0.04) (0.04)
Low-educa ed
Age
Con ol mean 26.05 27.26 27.38 28.02 29.79 30.66 31.73 33.31 34.44
Di e ence T-C −0.62 −0.65 0.57 1.20 1.17 0.73 0.81 −0.20 1.15
S anda d E o (0.91) (0.95) (0.89) (0.90) (0.90) (0.98) (0.96) (1.08) (0.99)
Ma ied
Con ol mean 0.59 0.66 0.58 0.60 0.63 0.60 0.61 0.67 0.67
Di e ence T-C 0.02 −0.10 −0.04 0.02 0.09 0.01 0.07 0.10 −0.01
S anda d E o (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.08) (0.07)
Sou ce: RDC o he Fede al S a is ical O ice and S a is ical O ices o he L˜
A¤nde , Mic ocensus 2008-2016; ⋆: The con ol
mean equals he a e age wo king hou s o women who ga e bi h o hei i s child in he las hal o 2007.; T ea men dummy
equals 1 i he i s child is bo n in he i s hal o 2007. Signi icance le els: * p<0.05, ** p<0.01, *** p<0.001. Robus
s anda d e o s in pa en heses.
18
Posi i e ea nings e ec s esul ing om he pa en al lea e e o m could u he be he conse-
quence o di e ences in he amily se ing caused by he e o m. Fo example, he new pa en al
lea e scheme could ha e a ec ed subsequen e ili y pa e ns and, he eby, indi ec ly mo he s’
ea nings in he long un. I he pa en al lea e e o m would ha e dec eased he p obabili y o
ha e ano he child, his could explain posi i e ea nings in he long un. Howe e , he li e a-
u e inds he opposi e: Cygan-Rehm (2016) in es iga ed whe he he e o m had an e ec on
iming o highe -o de bi hs. She inds ha high-income mo he s ha e a highe p obabili y o
a nex child wi hin 24 mon hs a e a p e ious childbi h. Mo eo e , Rau e (2019) compa es
e ili y a es p e- and pos - e o m and inds ha he e o m inc eased he p obabili y o gi e
bi h, in pa icula o mo he s who al eady ha e one o wo child en. On he o he , Klu e and
Schmi z (2018) ind no e ec s on he likelihood o ha e a subsequen bi h o mo he s wi h
high p io - o-bi h income.14 In conclusion, po en ial inc eases in e ili y would - i a all - lead
o a downwa d bias o ou es ima ion o he e o m e ec on daily ea nings se e al yea s a e
gi ing bi h o he i s child.
5.5 Changes in Employe S abili y
P e ious esea ch has shown ha as a esul o pa en al lea e e o ms, employe s abili y has
inc eased. Fo example, s udies by Bake and Milligan (2008) and Baum and Ruhm (2016)
ha e shown ha in cases whe e he employmen -p o ec ed pe iod o lea e (paid o unpaid)
was in oduced, employe s abili y has inc eased sign ican ly. Klu e and Schmi z (2018) ha e
shown ha e en o he Ge man 2007 e o m, whe e he job-p o ec ion pe iod o h ee yea s has
been le unchanged while only he du a ion and le el o pa en al lea e bene i paymen s ha e
chanced, employe s abili y has inc eased. In hei analysis based on he Mic ocensus, hey ind
ha mo he s ha e a highe p obabili y o e u n o hei p e-bi h employe . Ou analysis o
he IEB da a shows e y simila esul s. We ind ha high-income mo he s who e u n o wo k
in he second yea a e gi ing bi h ha e a 2 pe cen age poin s highe p obabili y o e u ning
o hei p e-bi h employe (Table 7).
Klu e and Schmi z (2018) a gue ha his inc eased employe s abili y is ewa ded by a
highe job-quali y in e ms o leng h o con ac . Simila ly, i could be ha employe s also
ewa d highe job s abili y wi h highe wages. The e o e, we compa e he long- un e ec s on
ea nings o mo he s who e u n o hei p e-bi h employe 15 wi h hose who e u n o he labo
ma ke wi h a new employe . As Table 8shows, he ea nings e ec s a e e y simila in bo h
g oups, in pa icula o high-income mo he s. No e ha bo h, he decision o e u n o he
labo ma ke as well as he decision o change he employe a e endogenous, and he e o e ou
14The au ho s ind on a e age a lowe p obabili y o ha e a subsequen bi h. In he Online Appendix, hey
analyze he p obabili ies o di e en sub-samples (e.g. by age o p io - o-bi h income).
15We e e o he employe as he es ablishmen . No e ha one company can ha e se e al es ablishmen s.
Thus, we would also iden i y a change o employe i someone mo es o ano he es ablishmen wi hin he same
company.
19
Table 7: E ec s o he pa en al lea e e o m on he p obabili y o change he i m a e he
employmen in e up ion
Low-income mo he s High-income mo he s
P obabili y o change he
employe
Con ol mean1Di e ence T-C Con ol mean1Di e ence T-C
y +1 0.041 −0.002 0.022 −0.005
(0.009) (0.006)
y +2 0.190 −0.016 0.096 −0.020**
(0.012) (0.007)
y +3 0.276 0.006 0.137 −0.006
(0.008) (0.013)
y +4 0.288 −0.003 0.151 −0.001
(0.012) (0.009)
y +5 0.299 −0.023 0.147 −0.005
(0.012) (0.008)
y +6 0.276 0.010 0.149 0.002
(0.012) (0.008)
y +7 0.254 0.003 0.131 −0.002
(0.011) (0.008)
y +8 0.244 0.002 0.113 0.003
(0.011) (0.007)
y +9 0.254 −0.004 0.119 −0.006
(0.011) (0.007)
Sou ce: IEB 1976-2016; own calcula ions. All speci ica ions con ol o seasonal ends (p e- e o m
dummy), 1 Con ol mean e e s o he a e age mean o mo he s who ga e bi h in he las qua e
o 2006, as he exponen ial o he log wage; The numbe o obse a ions a y be ween 6,358 (y +1) and
23,093 (y −1) o low-income mo he s and 8,128 (y +1) and 34,751 (y −1) o high-income mo he s. Sig-
ni icance le els: * p<0.05, ** p<0.01, *** p<0.001. Robus s anda d e o s in pa en heses.
20
esul s migh be d i en by selec ion e ec s. Howe e , he ac ha he ea nings e ec s o he
pa en al lea e e o m do no signi ican ly di e be ween mo he s who changed he employe a e
e u ning o he labo ma ke and hose who s ayed wi h he p e ious employe sugges s ha
he inc ease in job s abili y is p obably no he main d i e o he posi i e e ec s on mo he s’
ea nings.
Table 8: Di e ences in he e ec s o he pa en al lea e e o m on daily ea nings be ween mo he s
who e u ned o a new employe and hose who s ayed wi h he p e ious employe , o mo he s
wi h high and low p io - o-bi h ea nings
Low-income mo he s High-income mo he s
Daily
ea nings in
Eu o
Con ol Di e ence Di e ence Con ol Di e ence Di e ence
mean1Change -S aye Change -S aye mean1Change -S aye Change -S aye
wi h con ols2wi h con ols2
y -2 18.41 −0.503 −0.282 100.38 −1.051 0.794
(0.856) (0.726) (1.761) (1.349)
y -1 20.84 0.089 0.330 110.20 1.611 1.456
(0.723) (0.611) (1.542) (1.107)
y +1 5.72 1.795*** 1.563*** 15.94 7.403*** 3.641**
(0.472) (0.421) (1.801) (1.378)
y +2 15.51 −0.863 −0.442 59.09 −1.955 −3.027*
(0.701) (0.615) (2.191) (1.506)
y +3 17.93 −0.010 −0.482 59.32 0.092 0.193
(0.721) (0.825) (2.238) (1.555)
y +4 20.77 −0.166 −0.195 61.30 0.295 −0.002
(0.906) (0.792) (2.279) (1.515)
y +5 21.99 −1.716 −1.516 64.34 −3.253 −1.353
(0.970) (0.853) (2.271) (1.520)
y +6 22.84 −1.622 −1.174 67.04 0.074 0.779
(1.022) (1.022) (0.901) (1.526)
y +7 23.83 −2.501* −1.734 70.64 −2.038 −0.122
(1.075) (0.947) (2.203) (1.487)
y +8 24.00 0.027 0.439 73.67 −0.024 −0.167
(1.122) (0.988) (2.218) (1.510)
y +9 27.49 0.277 1.206 80.21 −1.080 0.157
(1.223) (1.076) (2.348) (1.592)
Sou ce: IEB 1976-2016; own calcula ions. All speci ica ions con ol o seasonal ends (p e- e o m dummy), 1 Con ol
mean e e s o he a e age mean o mo he s who ga e bi h in he las qua e o 2006, as he exponen ial o he log
wage; Con ols 2: P e-bi h wage, age a bi h, educa ion, expe ience ( & p ), el. du a ion o unemploymen , size o
es ablishmen , wo king ime be o e bi h, change o es ablishmen , eas Ge many, ci izenship, no. o child en, egion,
enu e and change o employe a e bi h. The numbe o obse a ions a y be ween 7,895 (y +1) and 29,179 (y −1)
o low-income mo he s and 9,283 (y +1) and 39,080 (y −1) o high-income mo he s. Signi icance le els: * p<0.05, **
p<0.01, *** p<0.001. Robus s anda d e o s in pa en heses.
21
5.6 Changes in Fa he s’ Pa en al Lea e Taking
Ano he po en ial mechanism ha could a ec mo he s’ ea nings is he ole o a he s. The
pe iod a e he bi h o he i s child is a c ucial ime o he di ision o household wo k
(e.g. Ya o sky, Kamp Dusch, and Schoppe-Sulli an, 2017). Pa e ni y lea e p omo es a he s’
in ol emen in childca e (and po en ially housewo k) and he eby has he po en ial o educe
sex specializa ion o ma ke wo k and unpaid ca e wo k and inc ease he mo he s’ ba gaining
powe . I he e is habi pe sis ence and lea ning in his con ex , hese e ec s could e en las
a e pa en al lea e has expi ed (e.g. Pa naik, 2019).
As desc ibed in sec ions 1and 2, he 2007 pa en al lea e e o m no only changed he amoun
and he du a ion o he pa en al lea e bene i bu also in oduced a ”daddy quo a”, i.e. a pe iod
o wo (ou o 14) mon hs o paid pa en al lea e ea ma ked o each pa en . This policy has
s ongly changed he ake-up o pa en al lea e by a he s. Whe eas less han h ee pe cen o
a he s wi h child en bo n be o e 2007 ha e aken pa en al lea e, his numbe has inc eased o 15
pe cen immedia ely a e he e o m and has been inc easing e e since (Sam leben, Schaepe ,
and W ohlich, 2019). Empi ical s udies ha e shown ha he sha e o a he s aking pa en al
lea e (while he mo he has e u ned o he labo ma ke ) has pa icula ly inc eased o a he s
in couples whe e bo h spouses ha e a uni e si y deg ee (Geisle and K eyen eld, 2019) and in
couples whe e he woman belongs o he highes ea nings qua ile (T appe, 2013). Ou analysis
based on he SOEP also shows ha a he s whose pa ne has a uni e si y deg ee ha e eac ed
much mo e s ongly o he e o m. The sha e o a he s aking pa en al lea e in his g oup has
isen om ze o in 2006 o mo e han wen y pe cen in 2007 (see Figu e 5).
Figu e 5: Sha e o Fa he s in Pa en al Lea e by Educa ional Deg ee o he Mo he
Sou ce: SOEP wa es 2010-2016; sha e o a he s wi h a spouse and who
had a child be ween 2000 and 2015. Own calcula ion.
22
P e ious esea ch om o he coun ies has shown ha a he s’ pa en al lea e aking causally
inc eases mo he s’ wages (e.g. D uedahl, E njaes, and Jo gensen, 2019), (e.g. Ande sen, 2018).
Fo example, D uedahl e al. (2019) show ha he in oduc ion o a pe iod o ea -ma ked
pa en al lea e o a he s in Denma k in 1998 inc eased he emale pa ne ’s sha e o household
income by 1.2 pe cen age poin s on a e age. This e ec is mos p onounced in he i s wo
yea s a e childbi h and diminishes o e ime, howe e emains signi ican un il eigh yea s
a e bi h. The au ho s can only specula e abou he mechanisms, bu a gue ha dec eased
gende speci ic specializa ion in he household can ha e a posi i e e ec on mo he s’ wages
(Ande sen, 2018). Ac ually, se e al s udies show ha a he s’ pa en al lea e aking causally
inc eases hei ime de o ed o childca e and housewo k e en in he pe iod a e hei pa en al
lea e. Exploi ing a pa en al lea e e o m in he Canadian p o ince o Quebec, Pa naik (2019)
shows ha a he s’ lea e aking inc eases hei in ol men in childca e and housewo k in la e
yea s. Simila esul s ha e also been ound o Ge many. Fo example, Tamm (2019) shows ha
e en sho pe iods o a he s’ pa en al lea e ha e long-las ing e ec s on a he s’ in ol emen
in childca e and housewo k. Simila ly, Schobe (2014) inds ha a he s’ pa en al lea e aking
causally inc eases hei in ol emen in childca e in he i s h ee yea s a e childbi h.
In he ligh o hese p e ious indings, we hypo hesize ha he inc ease in he ake-up o
pa en al lea e by a he s a e he 2007 pa en al lea e e o m in Ge many could ha e inc eased
a he s’ in ol emen in child ca e and household wo k and acili a ed mo he s’ e-en y o he
labo ma ke . The mo e equal di ision o unpaid ca e wo k could ha e inc eased mo he s’ labo
ma ke p oduc i i y, which in u n could be e lec ed in highe ea nings. Addi ionally, his could
explain he diminishing posi i e e ec s on ea nings o e ime: The e ec s a e s onges in he
i s yea s a e e-en e ing he labo ma ke bu ade ou as he child ge s olde .
6 Discussion and Conclusion
The Ge man pa en al lea e e o m o 2007, which inc eased he gene osi y o he inancial
bene i s o mo he s wi h medium and high p e-bi h ea nings bu cu he maximum du a ion
pe iod, has p olonged employmen in e up ions o high-income mo he s by h ee mon hs on
a e age. Ou es ima ion esul s show ha hese longe employmen in e up ions did no ha e
a nega i e e ec on mo he s’ long- e m ea nings pe spec i es. On he con a y, we e en ind
posi i e e ec s on mo he s’ ea nings, which diminish o e ime: In he sho un, mo he s wi h
high p e-bi h ea nings ea n en pe cen mo e as a esul om he new pa en al lea e scheme.
In he medium un, i.e. h ee o six yea s a e gi ing bi h, hey s ill ea n be ween 4 o 5
pe cen and in he long un, i.e. se en o nine yea s a e childbi h, 3 pe cen mo e han he
con ol g oups. This is also ue, albei o a lesse ex en , o mo he s wi h medium p e-bi h
ea nings. Fo mo he s wi h low p e-bi h ea nings, howe e , he e o m did no a ec ea nings
a e gi ing bi h.
23
