Tang, Can; Zhao, Zhong
Wo king Pape
In o mal ins i u ion mee s child de elopmen : Clan cul u e
and child labo in China
UNU-MERIT Wo king Pape s, No. 2022-032
P o ided in Coope a ion wi h:
Maas ich Economic and Social Resea ch Ins i u e on Inno a ion and Technology (UNU-MERIT),
Uni ed Na ions Uni e si y (UNU)
Sugges ed Ci a ion: Tang, Can; Zhao, Zhong (2022) : In o mal ins i u ion mee s child de elopmen :
Clan cul u e and child labo in China, UNU-MERIT Wo king Pape s, No. 2022-032, Uni ed Na ions
Uni e si y (UNU), Maas ich Economic and Social Resea ch Ins i u e on Inno a ion and Technology
(UNU-MERIT), Maas ich
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/326841
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-sa/4.0/









#2022-032

In o malins i u ionmee schildde elopmen :Clancul u eand
childlabo inChina


CanTangandZhongZhao










Published10Oc obe 2022






Maas ich EconomicandsocialResea chins i u eonInno a ionandTechnology(UNU‐MERIT)
email:[email p o ec ed]u|websi e:h p://www.me i .unu.edu

Boschs aa 24,6211AXMaas ich ,TheNe he lands
Tel:(31)(43)3884400

UNU-MERIT Wo king Pape s
ISSN 1871-9872
Maas ich Economic and social Resea ch Ins i u e on Inno a ion and Technology
UNU-MERIT
UNU-MERIT Wo king Pape s in end o dissemina e p elimina y esul s o esea ch ca ied
ou a UNU-MERIT o s imula e discussion on he issues aised.


In o mal Ins i u ion Mee s Child De elopmen :
Clan Cul u e and Child Labo in China
Can Tang
School o Business, Shanghai Uni e si y o In e na ional Business and Economics
Zhong Zhao※
School o Labo and Human Resou ces, Renmin Uni e si y o China
Augus , 2022
Abs ac
Using a na ional ep esen a i e sample, he China Family Panel S udies, his
pape explo es he in luences o clan cul u e, a hallma k o Chinese cul u al his o y, on
he p e alence o child labo in China. We ind ha clan cul u e signi ican ly educes
he incidence o child labo and wo king hou s o child labo e . The esul s exhibi
s ong boy bias, and a e d i en by boys a he han gi ls, which e lec s he pa ilineal
na u e o Chinese clan cul u e. Mo eo e , he impac is g ea e on boys om households
wi h lowe socioeconomic s a us, and in u al a eas. Clan cul u e ac s as a supplemen
o o mal ins i u ions: educes he incidence o child labo h ough isk sha ing and
easing c edi cons ain s, and helps o m social no ms o p omo e human capi al
in es men . We also employ an ins umen a iable app oach and ca y ou a se ies o
obus ness checks o u he con i m he indings.
Keywo ds: In o mal ins i u ion; Clan cul u e; Child labo ; China
JEL codes: J22, J81, O15
 We a e g a e ul o e y insigh ul discussions wi h and commen s om E ic V. Edmonds and Douglas
S aige , and would like o hank cons uc i e commen s om he edi o , wo anonymous e e ees, and
con e ence pa icipan s o he 7 h In e na ional Wo kshop on Economic Analysis o Ins i u ions o ganized
by Xiamen Uni e si y and semina pa icipan s a Xiamen Uni e si y, Cen al Uni e si y o Finance and
Economics, Kobe Uni e si y and Shanghai Uni e si y o Finance and Economics o hei help ul
commen s. We also hank he Ins i u e o Social Su ey a Peking Uni e si y o p o iding he China
Family Panel S udies used in his pape . Can Tang acknowledges inancial suppo om he Na u al
Science Founda ion o China (G an No. 72103130). Zhong Zhao acknowledges suppo by “Thema ic
Resea ch P ojec on China's Income Dis ibu ion” (P ojec No. 21XNLG03) wi h unding om he
esea ch und o Renmin Uni e si y o China.
※ Co esponding au ho : Zhong Zhao, m [email protected]. Zhong Zhao is also a membe o
GLO.
1
1 In oduc ion
Child labo has many sho - and long- e m nega i e e ec s, such as educ ion in
human capi al accumula ion (Heady, 2003; Gunna sson e al., 2006; Dumas, 2012;
O’Donnell e al., 2005) and ad e se e ec s on he u u e labo ma ke ou comes
(Eme son and Souza, 2011). Ob iously, he wel a e a bo h indi idual le el and
na ional le el will be impeded by such nega i e impac s.
Unde s anding he ac o s behind child labo is an impo an opic in he li e a u e.
In his pape , we ocus on he ole o clan cul u e on child labo in China. Ou s udy
shows ha clan cul u e ac s as a supplemen o o mal inancial ma ke by educing he
incidence o child labo h ough isk sha ing and easing c edi cons ain s, which
en iches he li e a u e on he mac o ac o s behind he incidence o child labo , such as
he ole o public policies, globaliza ion and inancial de elopmen (Dehejia and Ga i,
2005; Edmonds and Pa cnika, 2005; Dinopoulos and Zhao, 2007; Tang e al., 2020;
Zhao e al., 2021).
An impe ec inancial ma ke in de eloping coun ies leads o high le els o child
labo engagemen o wo easons (Baland and Robinson, 2000; Ranjan, 2001; Ud y,
2003). Fi s ly, he bene i s o child labo come om he cu en wage ea ned by he
child and he educed cos o schooling. Howe e , he cos o child labo is he u u e
labo ea nings impeded by in ol ed in child labo . When acing c edi cons ain s, a
highe in e es a e in an impe ec inancial ma ke leads o a lowe p esen discoun ed
alue o he u u e labo ea nings. Thus, he bene i s would be highe ela i e o he
cos , esul ing in highe incidence o child labo (Ud y, 2003). Secondly, decisions
ega ding child labo and schooling a e gene ally made by pa en s, which aise he issue
o agen p oblems. Agen p oblems become e en mo e salien when hey occu wi hin
impe ec inancial ma ke s (Ud y, 2003). E en amongs al uis ic pa en s, he
p opo ion o child labo can p oduce a socially ine icien equilib ium, as impe ec
inancial ma ke s p e en pa en s om ully in e nalizing nega i e e ec s (Baland and
Robinson, 2000; Ud y, 2003). Besides, as a pa o he household’s sel -insu ance

2
s a egy, child labo supply will inc ease when household acing nega i e shocks,
especially when he insu ance ma ke is no well de eloped (Jacoby and Skou ias ,1996;
Beegle e al., 2006; De Jan y e al., 2006).
Cul u e is a i al in o mal ins i u ion, and is iden i ied in eme ging economies as
a c ucial de e minan o many economic ou comes, especially when he o mal
ins i u ion is poo ly de eloped (Guiso e al., 2006; Fe nández, 2011; Voig lände and
Vo h, 2012; Alesina and Giuliano, 2015). Howe e , he e a e ew economic s udies
ocusing on he impac o cul u e on child labo . Thus, ou s udy also con ibu es o he
g owing li e a u e o cul u al economics by examining he impac o clan cul u e on he
incidence o child labo in China.
We ocus on China o h ee easons. Fi s ly, al hough he en ollmen a es o
p ima y and junio high schools a e high in China
1
, app oxima ely 7.74% o child en
aged be ween 10 and 15 yea s we e engaged in child labo in 2010 (Tang e al., 2018).
I is c ucial o in es iga e how o educe he child labo pa icipa ion in China. Secondly,
clan cul u e is a well-known hallma k o Chinese cul u e (G ei and Tabellini, 2010,
2017; Zhang, 2020). Whils o icial clan o ganiza ions we e la gely abolished ollowing
he Chinese Communis Pa y’s ise o powe in 1949, he in luences o clan cul u e a e
likely o pe sis and con inue o ha e an impo an impac on he beha io o indi iduals
in mode n China (G ei and Tabellini, 2010, 2017). Thi dly, China is he la ges
ansi ion and de eloping economy, bu has ye o es ablish a well-de eloped inancial
ma ke , legal sys em, and social sa e y sys em, pa icula ly in u al a eas. In such
con ex , clan cul u e plays he ole o o mal ins i u ions in p omo ing in o ma ion and
isk sha ing, educing ansac ion cos s, o e coming inancial cons ain s o
en e p ises, and easing c edi cons ain s o i s membe s (Webe , 1981; Peng, 2004;
G ei and Tabellini, 2010; Cox and Fa champs, 2007; Zhang, 2020). Besides, clans also
p o ide public goods and se ices o clan membe s (Fei, 1946; F eedman, 1965; Tsai,
2007; Ge a d e al., 2015).
1
The en ollmen a e o child en aged be ween 10 o 15 yea s is 97% in 2010, acco ding o ou
calcula ion based on CFPS 2010.
3
Combing he heo ies o child labo de e mina ion and he unc ions o clan cul u e,
we hypo hesize ha clan cul u e can in luence he engagemen o child labo . Clan
cul u e, as a supplemen o o mal ins i u ions, may educe he labo supply o child en
om poo amilies by easing c edi cons ain s and p omo ing isk sha ing. Mo eo e ,
his impac is supposed o be biased owa d boys due o he pa ilineal clan na u e.
To es ou hypo heses, we use he indica o o whe he a household has a
genealogy book as a key measu e o he s eng h o clan cul u e a he household le el
ollowing exis ing li e a u e (Peng, 2004; Chen e al., 2016; G ei and Tabellini, 2017;
Zhang, 2020). Ha ing a genealogy book means he household subsc ibing o a clan.
Thus, his pape ocus on he di e ences be ween clan membe s and non-clan membe s,
o he han he impac s o clan cul u e o egional ou comes o all indi iduals. Tha is,
a he han explo ing he ole o ex e nal cul u al en i onmen (Guiso e al., 2004), we
pay mo e a en ion o he in e gene a ional cul u al ansmission o p e e ences and
belie s wi hin amilies (Tabellini, 2008), social no m among clan membe s, and he
e ec o being a clan membe .
Besides using a household le el indica o , we also explo e o measu e he s eng h
o clan cul u e by he su name concen a ion in he illage ollowing Peng (2004).
Employing ou wo measu es o clan cul u e, we a emp o sepa a ely iden i y he
po en ial e ec s o ex e nal o igins and amily o igins. On he one hand, concen a ion
o he same su name e lec s he s eng h o clan cul u e in he illage, ega dless o
household no ms. On he o he hand, he genealogy book measu es he impo ance o
clan cul u e o a pa icula household, a e con olling he illage le el indica o o clan
cul u e.
Based on na ional ep esen a i e da a, he China Family Panel S udies, we ind
ha clan cul u e can signi ican ly educe he incidence o child labo . The esul s a e
d i en by boys a he han gi ls. The di e en ial e ec s be ween boys and gi ls indica e
a s ong eng ained bias owa ds boys wi hin he clan cul u e since i s pa ilineal na u e,
which in luences pa en s o alloca e esou ces owa ds boys. Mo eo e , he impac is
4
g ea e on boys om households wi h lowe socioeconomic s a us, and in u al a eas.
We in es iga e isk sha ing, in o mal inancial ins i u ion, social no m on
p omo ing educa ion h ee mechanisms h ough which clan cul u e in luences he
incidence o child labo . Speci ically, clan cul u e has a g ea e impac on educing
child labo engagemen amongs boys when se ious disas e s occu . Addi ionally, clans
a e mo e likely o help a he s o boys, i hey all in o di icul y. Finally, we p o ide
e idence o how clan cul u e may inc ease s udy ime and household educa ional
expendi u e on boys.
Ou pape con ibu es o exis ing li e a u e in se e al ways. Fi s ly, i en iches he
li e a u e o explo ing he impac s o cul u e on economic de elopmen . As exis ing
li e a u e shows, cul u e can e ec i ely in luence labo pa icipa ion decision, e ili y
decision, schooling decisions, academic pe o mance, wage a e, and labo mobili y
(e.g., Fe nández, 2007; Guiso e al, 2008; Fe nández and Fogli, 2009; Gia azzi e al.,
2013; Munshi and Rosenzweig, 2006; Pi , Rosenzweig and Hassan, 2012). Howe e ,
seldomly do hey ocus on child labo pa icipa ion. We emphasize ha cul u e can also
in luence human capi al accumula ion and economic de elopmen h ough he child
labo channel.
Secondly, we con ibu e o exis ing child labo de e mina ion li e a u e (Dehejia
and Ga i, 2005; Edmonds and Pa cnika, 2005; Basu e al., 2010; Tang e al., 2018;
Tang e al. 2020; Zhao e al., 2021) by iden i ying a new in luencing ac o : clan cul u e.
Since o mal ins i u ions a e weak in mos de eloping coun ies, explo ing he ole o
cul u e in child labo p o ides a new insigh o unde s anding he de e mina ion and
dis ibu ion o child labo .
Thi dly, his pape highligh s he po en ial h ea associa ed wi h eliance on
in o mal ins i u ions and cul u e, such as clans, o ill he oid o unde -de eloped
o mal ins i u ions, since he in o mal ins i u ion and cul u e can be gende biased.
The emainde o he pape is o ganized as ollows: Sec ion 2 b ie ly in oduces
clan cul u e in China; Sec ion 3 desc ibes he da a and ou lines he empi ical s a egy;
5
Sec ion 4 p esen s and discusses he main esul s; Sec ion 5 p o ides obus ness checks;
Sec ion 6 conduc s he e ogenei y analyses; Sec ion 7 explo es po en ial mechanisms
and possible in e p e a ions o he main indings; Sec ion 8 p o ides concluding
commen s.
2 Clan Cul u e in China and Rela ed Li e a u e
Clan cul u e is a hallma k cul u e in China (G ei and Tabellini, 2010, 2017; Zhang,
2020). The clans a e kinship-based pa ilineal o ganiza ions made up o indi iduals
wi h one common male ances o . The ea lies clan o ganiza ions appea ed wi hin he
Zhou Dynas y (1046-256 BC) and hey p e ailed wi hin he Song Dynas y (960-1279
AD) when la ge con lic s b oke ou equen ly in no he n China, which led o g ea
popula ion mig a ion o he sou h; he p esen spa ial dis ibu ion o clans was mainly
o med a ha ime (Feng, 2013). Zhu Xi (1130–1200) was a e en ad oca e o clans
and p o ided de ailed ins i u ional designs o Chinese clan o ganiza ions, which
became he s anda d social p ac ices in he subsequen eigh cen u ies (Cheng e al.,
2021). Fo managemen con enience and s onge cohesi eness, a adi ional clan
ypically had an elde boa d o adminis e clan a ai s, an ances al emple o o e
sac i ices o ances o s, a genealogy book eco ding b ie biog aphies o all descen s
om he apical ances o , and a code o conduc o di ec and cons ain clan membe s.
Besides, a ypical clan usually has common p ope ies ha can be used o assis
membe s in need.
Chinese clans ha e wo basic ea u es. Fi s ly, clans app ecia e amily alues,
obliga ions and loyal y (Fei, 1946; Liu, 1959, Peng, 2004). Due o he sense o du y,
clan membe s a e mo e inclined o make clan-o ien ed decisions. In e u n, clans also
p o ide p o ec ion o loyal membe s. Howe e , clan cul u e leads o s ong boy
p e e ence. As coope a ion o ganiza ions consis ing o pa ilineal households ha ace
hei o igin o a common male ances o , clans emphasize pa ilineal descen so ha only
males can pass on hei amily names o nex gene a ions and ha e he igh o be clan
leade s (Feng, 2013). Along wi h clan cul u e, he Chinese amily sys em is pa ia chal,
12
All models a e es ima ed using he sample weigh s p o ided by he CFPS. In all
eg essions, s anda d e o s a e adjus ed o clus e ing a he household le el o accoun
o any po en ial co ela ion wi hin he same household (Be and e al., 2004)
9
.
Al hough we only ha e c oss-sec ional da a, cul u e is usually ega ded as
exogenous. The cul u al ai s ansmi om gene a ion o gene a ion (G ie , 1994;
Guiso e al., 2006), and in luence economic ou comes h ough wo po en ial channels:
belie s (i.e., p io s) and alues (i.e., p e e ences), which can only be eshaped in he
long e m (Guiso e al., 2006). Thus, cul u e can be ega ded as pe sis en and s able.
Acco ding o Becke (1996, p. 16): “Indi iduals ha e less con ol o e hei cul u e han
o e o he social capi al. They canno al e hei e hnici y, ace o amily his o y, and
only wi h di icul y can hey change hei coun y o eligion. Because o he di icul y
o changing cul u e and i s low dep ecia ion a e, cul u e is la gely a ‘gi en’ o
indi iduals h oughou hei li e imes.”
Fu he mo e, whe he a household holds a genealogy book is ela i ely exogenous
o he household. A genealogy book usually eco ds he his o y o a clan, including b ie
biog aphies o all descen s om he apical ances o and desc ip ions o all mega e en s
and hono able descen s in he clan. Compiling a genealogy equi es de ailed
in o ma ion on all ex ended- amily membe s including p e ious gene a ions and
cu en gene a ion. The e o e, whe he a household holds a genealogy book is mos ly
de e mined by he ex ended- amily o he han an indi idual household.
4 Main Resul s
In his sec ion, we i s ly p esen esul s om he baseline model, and hen deal
wi h he po en ial endogenei y issues. The main ou comes o in e es a e child labo
dummy a iable and sum o paid and unpaid o wo king hou s o child en.
9
We also conside clus e ing s anda d e o a he illage/communi y le el as obus ness checks. The
main empi ical esul s a e s ill hold.

13
4.1 Baseline esul s
Table 4 epo s he baseline esul s. All eg essions in Table 4 include indi idual
cha ac e is ics, household cha ac e is ics, illage/communi y cha ac e is ics, and
coun y ixed e ec s. The es ima ed coe icien s o clan cul u e s eng h a e nega i e
and s a is ically signi ican in Column (1), sugges ing ha clan cul u e helps comba
child labo . Pa icula ly, a child om a clan household is abou 2.7% less likely o be a
child labo e .
Since clans in China emphasize pa ilineal descen , clan cul u e leads o a s ong
boy p e e ence. Thus, we u he explo e whe he clan cul u e has di e en ial e ec s
on boys and gi ls. The esul s a e shown in Columns (2)-(4) in Table 4. We ind ha all
he e ec s go h ough he in e ac ion e m o genealogy and boy; he coe icien s o
genealogy become insigni ican in Column (2). In o he wo ds, clan cul u e helps
comba child labo only o boys. We un sepa a e eg essions o quan i y he
di e en ial impac on boys and gi ls, wi h he esul s o gi ls in Column (3) and boys
in Column (4). Resul s show ha clan cul u e signi ican ly helps educe he incidence
o child labo o boys. Howe e , i does no ha e a signi ican e ec o gi ls.
[Table 4 abou he e]
Since he e is no signi ican di e ence be ween he o e all likelihood o child
labo o boys and gi ls in China (Tang e al., 2018),
10
he s onge impac o clan
cul u e on boys may no come om he di e en le el o child labo incidence be ween
boys and gi ls. When a amily is aced wi h se e e c edi cons ain s, bo h sons and
daugh e s may ha e o engage in economic ac i i ies. Clan cul u e may elax household
c edi cons ain s and lead o in a-household esou ce ealloca ion owa ds boys due o
s ong boy p e e ence wi hin clans (we will e i y his he eina e ). Consequen ly, a
elaxa ion o esou ce cons ain s by clans may c ea e di e en ial e ec s o boys and
gi ls.
10
In he sample used in ou analyses, 7.6% o gi ls and 7.9% o boys engaged in child labo .
14
Besides economic ac i i ies, child en also in ol e non-economic housewo k, such
as aking ca e o he amily and household cho es. We also examine he child en’s non-
economic housewo k pa icipa ion. Acco ding o ou calcula ion om he CFPS 2010,
51.34% o child en pa icipa e in housewo k. On a e age, child labo e s spend 6.72
hou s pe day on indi idual wo k on a e age, whils hey only spend 0.847 hou s pe
day on housewo k. When we conside he impac s o clan cul u e on child en’s
housewo k pa icipa ion, we ind no signi ican e ec s. In o he wo ds, clan cul u e
helps educe child en’s pa icipa ion in economic ac i i ies, bu has no impac s on hei
housewo k pa icipa ion.
11
4.2 Co ec ing o he po en ial endogenous bias
The endogenei y conce ns may come om h ee aspec s in his esea ch: e e se
causali y, measu emen e o o clan cul u e, and omi ed a iables.
Since whe he a household has a genealogy book is ha d o be in luenced by child
labo engagemen , e e se causali y issue is no he main conce n in ou speci ica ion.
As o measu emen e o o clan cul u e, i is possible o a household emphasizing
clan cul u e do no keep a genealogy book, bu seldomly would a non-clan household
keep a genealogy book. Measu emen e o in his di ec ion unde es ima es he impac
o clan cul u e. The e o e, he majo endogenei y conce ns may come om omi ed
a iables.
To assess he impo ance o po en ial omi ed a iable bias, we pe o m a es
p oposed by Os e (2019). This me hod analyzes he di e ences in coe icien s o
in e es and he R-squa ed o he main eg essions be ween excluding con ols and
including con ols. I including con ols inc eases he R-squa ed o he model while no
a ec ing he coe icien o in e es , i is less likely ha including unobse ables would
bias he esul s. Os e (2019) assesses his po en ial bias by compu e he ela i e
impo ance o unobse ables o obse ables (𝛿) ha would be consis en wi h a
11
The esul s a e a ailable om he au ho s upon eques .
15
coe icien o in e es equal o ze o (𝛽=0). Tha is, how impo an should he
unobse ables o be ela i e o he obse ables o elimina e he es ima ed e ec .
Fo ou baseline speci ica ion, he 𝛿 ha ma ches 𝛽=0 amoun s o 2.208 o whole
sample, 3.629 o gi ls and 1.871 o boys. I means ha he impo ance o
unobse ables would ha e o be 2.208, 3.629 and 1.871 imes highe han ha o he
obse ables o he coe icien s o be ze o. Acco ding o Os e (2019), he omi ed
a iable bias would be less o a conce n i he es ima ed 𝛿 exceeds one. The alues o
𝛿 a e a mo e han one in ou es s, indica ing ha omi ed a iables bias is limi ed.
To u he con ol o he possible endogenous issue and assess he impo ance o
he po en ial endogenous bias, we adop an ins umen a iable (IV) app oach and ca y
ou a Hausman es . Re e ing o Chen e al. (2021), we ins umen clan cul u e wi h a
ci y’s sho es dis ance o he academic cen e whe e Zhu Xi (1130–1200 CE)
de eloped and sp ead his philosophy.
Zhu Xi was a e en ad oca e o clans and p o ided de ailed ins i u ional designs
o Chinese clan o ganiza ions, which became he s anda d social p ac ices in he
subsequen eigh cen u ies (Cheng e al., 2021). Zhu had se up 3 academies du ing his
whole li e, all o which we e loca ed in Nanping (in Fujian p o ince) and he majo i y
o Zhu’s classical ex s we e comple ed he e, such as Jiali ( he Family Ri uals) (Chan,
1987). In Jiali, Zhu no only elabo a ed on he impo ance o es ablishing clan common
p ope ies, such as ances al halls, ances al g a eya ds, and i ual land, bu also laid
ou de ailed ins uc ions on he design o ances al halls and he ules o amily i uals
like adul ce emonies, ma iages, une als, and ances al o e ings e c. (Cheng e al.
2021). Meanwhile, o he h ee academies Zhu buil , he mos impo an one is “Kao ing
Academy”, because mos o his amous pupils a ended him he e om all o e China
and mos o he eco ded con e sa ions ook place he e, o ming he “Kao ing School”
wi h g ea in luence in his o y (Chan, 1987). In a wo d, Nanping was he academic
cen e o sp eading Zhu Xi’s main ideas abou clan cul u e. Thus, we expec ha egions
close o Nanping we e mo e exposed o Zhu Xi’s in luences his o ically and, hanks o
16
cul u al pe sis ence, a e mo e likely o p ese e s ong clan cul u e he i age oday.
Besides, he loca ion choice o he academic cen e whe e Zhu Xi (1130–1200 CE)
de eloped and sp ead his philosophy is la gely andom. Zhu Xi chose o es ablish
academies in Nanping because his a he used o be a local go e nmen o icial he e
and his mo he was bu ied he e. Nanping was nei he an economic cen e no backwa d
egions. The e o e, consis en wi h Chen e al. (2021) and Cheng e al. (2021), we a gue
ha he IV does no a ec he child labo decision h ough o he channels han clan
cul u e. So, i is easonable o use a ci y’s sho es dis ance o Nanping as an IV o he
s eng h o clan cul u e.
The ins umen al a iable (IV) esul s a e shown in he Table 5. These esul s
ein o ce ou baseline conclusion: he nega i e e ec s o clan cul u e on he incidence
o child labo a e signi ican o he ull sample and he boy sample.
[Table 5 abou he e]
Fu he mo e, we employ a Hausman speci ica ion es o examine whe he i is
necessa y o adop an IV app oach (Hausman, 1978). The esul om a obus Hausman
es using boo s ap wi h 1500 eplica ions shows P ob > chi2=0.23, which means ha
we canno ejec he null hypo hesis ha all explana o y a iables a e exogenous.
Since OLS is mo e e icien han IV app oach i all explana o y a iables a e
exogenous, and bo h es o Os e (2019) and Hausman es indica e ha he e is no
signi ican endogenous bias, i is a ional o us o choose OLS me hod as he main
empi ical s a egy in his pape
12
.
12
I should be added ha , in he OLS eg essions we con ol o he coun y ixed e ec . When i comes o he IV
app oach, since ou ins umen a iable - a ci y’s dis ance o Nanping - is a pa o he coun y ixed e ec , we canno
con ol o he coun y ixed e ec anymo e due o pe ec collinea i y. Thus, o execu e he Hausman speci ica ion
es , he eg ession equa ion o OLS we use he e is he same as he eg ession equa ion o IV app oach (wi hou
con olling o coun y ixed e ec ). Acco ding o Mundlak (1987), we can always ind a linea p ojec ion o ans e
a ixed e ec model o a andom e ec model, so i is a ional o ans e he OLS eg ession equa ion con olling
o coun y ixed e ec o an equa ion wi hou con olling o coun y ixed e ec .
17
4.3 Wo king hou s
In his pa , we go one s ep u he o examine he e ec s o clan cul u e on
wo king hou s o child en. Since a subs an ial p opo ion o child en a e obse ed wi h
ze o wo king hou , we use a obi model. Resul s in Table 6 show ha clan cul u e
signi ican ly helps educe child en’s wo king hou s, and again his esul is d i en by
boys o he han gi ls.
[Table 6 abou he e]
Because a majo i y o child en a e no wo king, i is impo an and in e es o
in es iga e how he clan cul u e e ec s on he ex ensi e ma gin ( he pa icipa ion o in
he labo ma ke ) and on he in ensi e ma gin (wo king hou s condi ional on wo king)
by McDonald and Mo i 's (1980) decomposi ion.
Table 7 epo s he decomposi ion esul s. Panel A and B show esul s o he whole
sample, while Panel C o gi ls and D o boys. Clan cul u e has educed bo h he
ex ensi e and he in ensi e ma gin, bu he e ec s a e again signi ican o boys only.
Panel D illus a es ha clan cul u e educes labo pa icipa ion o boy by 2.6% a
ex ensi e ma gin and educes wo king ime by 0.24 hou s a in ensi e ma gin.
[Table 7 abou he e]
5 Robus ness checks
In his subsec ion, we conduc a se ies o obus ness checks o ou baseline
esul s: pe mu a ion es ; sample es ic ion; al e na i e measu e o he s eng h o
clan cul u e; aking no h-sou h di e ence in o accoun ; and conside a ion o di e en
e hnic g oups.
5.1 Pe mu a ion es
We adop a pe mu a ion es o Rosenbaum (2007) o check whe he he
es ima ed esul s a e eally signi ican o jus due o andom chance. The pe mu a ion

18
es makes no assump ions on he unde lying dis ibu ion o he da a. In he
pe mu a ion es , he null hypo hesis is ha he clan cul u e has no e ec on he odds
o child labo . Unde he null hypo hesis, he es ima ed coe icien om he ac ual
da a can be conside ed a andom sample om he pe mu a ion dis ibu ion. We can
p oduce he pe mu a ion dis ibu ion o he es ima ed coe icien s and use i o
s a is ical in e ence. We andomly assign whe he o no he amily household has a
genealogy book as he placebo ea men s a us o each child (Rosenbaum, 2007; Lu
and Ande son, 2015). Speci ically, he a io o households ha ing genealogy books
in placebo ea men s a us is consis en wi h ha in he ac ual da a. We es ima e he
placebo ea men e ec on he incidence o child labo . The dis ibu ion o placebo
ea men e ec s om 4000 andom assignmen s a e displayed in Fig.2. The dashed
line shows he es ima ed ea men e ec om he baseline analysis. The p alue o
he pe mu a ion placebo es is he p opo ion o placebo es ima es ha a e equal o
o la ge in absolu e alue han he co esponding es ima e om he baseline analysis.
We ind ha he p alue is 0.033, which ejec s he null hypo hesis o no e ec , hus
p o iding u he suppo o ou iden i ica ion s a egy and main indings.
[Figu e 2 abou he e]
5.2 Sample es ic ion
Clans emphasize pa ilineal descen and ha e a s ong boy p e e ence. Thus, when
we di ide he whole sample in o he subsamples o boys and gi ls, he classi ica ion
may by endogenous. As amilies wi h sons may be mo e deeply in luenced by clan
cul u e, i.e., h ough e ili y decisions, i is possible ha he di e en ial gende e ec s
shown in he baseline esul s may no only cap u e he impac o clan cul u e on he
incidence o child labo , bu also include di e en s eng hs o clan cul u e be ween
hese wo subsamples. As a obus ness check, we es ic ou analysis o he child en
om households wi h bo h sons and daugh e s. On he one hand, such amilies a e mo e
likely o ha e he same gende p e e ence, bu on he o he hand, in a-household
esou ce alloca ion be ween boys and gi ls is mo e ele an o amilies wi h bo h sons
19
and daugh e s. The esul s a e in Table 8: clan cul u e s ill signi ican ly educes he
incidence o child labo o boys, bu no o gi ls, and he di e en ial gende e ec s
a e indeed s onge o such amilies.
13
[Table 8 abou he e]
5.3 Al e na i e measu es o he s eng h o clan cul u e
We hen check i ou esul s a e sensi i e o al e na i e measu es o clan cul u e
s eng h. Whe e clan cul u e is p e alen , ex ended amily membe s end o li e in he
same illage. In e e ence o Peng (2004), we measu e he s eng h o clan cul u e by
he su name concen a ion in he illage. Pa icula ly, as shown in Columns (1)-(3) in
Table 9, we adop whe he o no he e is a su name sha ed by o e 10% o illage
households as an al e na i e measu e o he s eng h o clan cul u e.
14
We exclude he
u ban sample in his pa , as we do no ha e any in o ma ion abou su name p e alence
a communi y le el wi hin u ban a eas. Using he al e na i e measu e o clan cul u e
s eng h does no al e ou main inding: clan cul u e has a nega i e and signi ican
e ec on he incidence o child labo o boys.
Fu he mo e, some li e a u e ocuses on he in e gene a ional cul u al ansmission
o p e e ences and belie s wi hin amilies (Tabellini, 2008), whils o he s shed ligh on
he ole o ex e nal cul u al en i onmen (Guiso e al., 2004). Thus, an in e es ing
ques ion o conside is whe he i is in e gene a ionally ansmi ed clan cul u e wi hin
amilies, o he a ea in which an indi idual li es ha helps comba child labo .
Employing ou wo measu es o clan cul u e, we a emp o sepa a ely iden i y he
po en ial e ec s o ex e nal o igins and amily o igins. On he one hand, concen a ion
o he same su name e lec s he s eng h o clan cul u e in he illage, ega dless o
household no ms. On he o he hand, he genealogy book measu es he impo ance o
13
We also conside he impac s o in e p o incial mig an s o ou esul s. Only 1.4% o ou sample a e
in e p o incial mig an s. When we es ic ou sample o local child en, ou main esul s do no change
much. The esul s a e a ailable om he au ho s upon eques .
14
This indica o comes om a ques ion o CFPS2010 in illage/communi y module, which ask “In you
illage, was he e any su name sha ed by o e 10% o all he households?”. The e is no addi ional
in o ma ion a ailable o use o he cu o .
20
clan cul u e o a pa icula household, a e con olling he illage le el indica o o clan
cul u e.
Speci ically, we add clan cul u e measu es o bo h illage le el and household
le el o he eg essions and epo he esul s in Columns (4)-(6) o Table 9. In e es ingly,
bo h indica o s a e signi ican ly nega i e in he eg ession o boys. Tha is, clan cul u e
does help educe child labo pa icipa ion o boys i espec i e o whe he i is
ansmi ed wi hin he amily o p e alen in he illage.
[Table 9 abou he e]
5.4 Conside di e ence in clan cul u e among e hnic g oups
The e a e 56 e hnic g oups in China, including he Han Chinese accoun ing o
abou 91.51% o he o al popula ion and es a e 55 o he e hnic mino i ies.
15
Since
di e en e hnic g oups may hold di e en belie s, we addi ionally check o obus ness
ac oss di e en e hnic g oups by con olling he in e ac ion o he s eng h o clan
cul u e and whe he he illage/ esiden ial communi y is a mino i y-concen a ed a ea.
The esul s a e epo ed in Table 10. The es ima ed esul s a e qui e close o hose
ob ained om baseline esul s. The coe icien s o he dummy a iable on whe he a
household belongs o Han Chinese, i a illage/communi y is a mino i y-concen a ed
a ea, and he in e ac ion e m o clan cul u e and mino i y-concen a ed a ea a e all
insigni ican , which implies ha clan cul u e plays an impo an ole in educing child
labo amongs bo h he Han Chinese and mino i ies.
[Table 10 abou he e]
5.5 Take no h-sou h di e ence in o accoun
Figu e 1 shows clan cul u e concen a es in sou he n egions o China. The
une en dis ibu ion may imply di e en impac s o clan cul u e on child labo
15
Da a sou ce: 2010 Na ional Census o China.
21
pa icipa ion be ween no he n and sou he n p o inces. As shown in Table 11,
columns (1)-(3) show he es ima es o child en li ing in no he n p o inces whils
he es ima es o child en li ing in sou he n p o inces a e epo ed in Columns (4)-
(6). We ind ha clan cul u e signi ican ly educes he incidence o child labo o
child en in sou he n p o inces, especially o boys, bu has no signi ican e ec on
he engagemen o child labo o child en in no he n p o inces.
16
[Table 11 abou he e]
6 Mechanisms and In e p e a ions
6.1 A supplemen o o mal ins i u ions
Since cul u e ac s as a supplemen o o mal ins i u ions, especially when he
o mal ins i u ion is no well de eloped (Alesina and Giuliano, 2015), he impac o
clan cul u e on child en’s labo o ce pa icipa ion may a y be ween u ban and u al
a eas due o he di e en de elopmen o o mal ins i u ion and economic le el. We
in es iga e u al-u ban he e ogenei y o es whe he clan cul u e ac s as a supplemen
o o mal ins i u ions and plays a much impo an ole in u al China.
As shown in Table 12, Panel A p esen s he e ec s o clan cul u e by u al/u ban
hukou s a us, while Panel B e e s o he e ec s o clan cul u e by u al/u ban
esidence. Columns (1)-(3) show he es ima es o u ban child en whils he es ima es
o u al child en wi h a e epo ed in Columns (4)-(6). The esul s show ha clan
cul u e helps educe he incidence o child labo o u al child en, especially o u al
boys, bu has no signi ican e ec on he likelihood o child labo amongs bo h u ban
gi ls and u ban boys.
[Table 12 abou he e]
16
No he n China includes Beijing, Tianjin, Hebei, Shanxi, Inne Mongolia, Liaoning, Jilin,
Heilongjiang, Shandong, Henan, Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang. Sou he n China
includes Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Jiangxi, Hubei, Hunan, Guangdong, Guangxi,
Hainan, Chongqing, Sichuan, Guizhou, Yunnan, and Tibe .
28
Cox, D., Fa champs, M., 2007. Chap e 58 Ex ended Family and Kinship Ne wo ks:
Economic Insigh s and E olu iona y Di ec ions. Else ie B.V.
de Jan y, A., Finan, F., Sadoule , E., Vakis, R., 2006. Can condi ional cash ans e s
se e as sa e y ne s in keeping child en a school and om wo king when
exposed o shocks? Jou nal o De elopmen Economics 79, 349–373.
Dehejia, Rajee H., Ga i, R., 2005. Child Labo : The Role o Financial De elopmen
and Income Va iabili y ac oss Coun ies. Economic De elopmen & Cul u al
Change 53, 913-931.
Dinopoulos, E., Zhao, L., 2007. Child Labo and Globaliza ion. Jou nal o Labo
Economics 25, 553-579.
Dumas, C., 2012. Does Wo k Impede Child Lea ning? The Case o Senegal.
Economic De elopmen & Cul u al Change 60, 773-793.
Edmonds, E. V., 2008. Child Labo . Handbook o De elopmen Economics, 4, 3607-
3709
Edmonds, E.V., Pa cnik, N., 2005. The e ec o ade libe aliza ion on child labo .
Jou nal o In e na ional Economics 65, 401-419.
Eme son, P., Souza, A., 2007. Child labo , school a endance and in a-household
gende bias in B azil. Wo ld Bank Economic Re iew 21, 301–316.
Eme son, P.M., Souza, A.P., 2011. Is Child Labo Ha m ul? The Impac o Wo king
Ea lie in Li e on Adul Ea nings. Economic De elopmen and Cul u al Change
59, 345-385.
Fei, H.T., 1946. Peasan y and Gen y: An In e p e a ion o Chinese Social S uc u e
and I s Changes. Ame ican Jou nal o Sociology 52, 1-17.
Feng, E., 2013. Zhongguo Gudai de Zongzu he Ci ang [clans and Ance al emples in
ancien China]. The Comme cial P ess.
Fe nández, R., 2007. Al ed Ma shall Lec u e: Women, Wo k, and Cul u e. Jou nal
o he Eu opean Economic Associa ion 5, 305-332.
Fe nández, R., 2011. Chap e 11 - Does Cul u e Ma e ? in: Benhabib, J., Bisin, A.,
Jackson, M.O. (Eds.), Handbook o Social Economics, Vol. 1. No h-Holland,
pp. 481-510.
Fe nández, R., Fogli, A., 2009. Cul u e: An Empi ical In es iga ion o Belie s, Wo k,
and Fe ili y. Ame ican Economic Jou nal Mac oeconomics 1, 146-177.
F eedman, B.M., 1965. Lineage o ganiza ion in sou heas e n China. Paci ic A ai s
32, 207.
Ge a d, P.I.M., Qian, N., Xu, Y., Yao, Y., 2015. Making Democ acy Wo k: Cul u e,
Social Capi al and Elec ions in China. Social Science Elec onic Publishing.

29
Gia azzi, F., Schian a elli, FM., Se a inelli, MF., 2013. A i udes, Policies, and Wo k
a i udes, polices, and wo k. Jou nal o he Eu opean Economic Associa ion 11,
1256–1289.
G ei , A., 1994. Cul u al Belie s and he O ganiza ion o Socie y: A His o ical and
Theo e ical Re lec ion on Collec i is and Indi idualis Socie ies. Jou nal o
Poli ical Economy 102, 912-950.
G ei , A., Tabellini, G., 2010. Cul u al and Ins i u ional Bi u ca ion: China and
Eu ope Compa ed. Ame ican Economic Re iew 100, 135-140.
G ei , A., Tabellini, G., 2017. The clan and he co po a ion: Sus aining coope a ion
in China and Eu ope. Jou nal o Compa a i e Economics 45, 1-35.
Guiso, L., Mon e, F., Sapienza, P., Zingales, L., 2008. Di e si y. Cul u e, gende , and
ma h. Science 320, 1164.
Guiso, L., Sapienza, P., Zingales, L., 2004. The Role o Social Capi al in Financial
De elopmen . Ame ican Economic Re iew 94, 526-556.
Guiso, L., Sapienza, P., Zingales, L., 2006. Does Cul u e A ec Economic Ou comes?
Jou nal o Economic Pe spec i es 20, 23-48.
Gunna sson, V., O azem, P.F., Sánchez, M.A., 2006. Child Labo and School
Achie emen in La in Ame ica. Wo ld Bank Economic Re iew 20, 31-54.
Ha ell, S., 2002. Pa iliny, pa ia chy, pa imony su ace ea u es and deep s uc u es
in he Chinese amily sys em. Pape d a . URL
h p:// acul y.washing on.edu/s e eha /PPP.h ml
Hausman, J. A., 1978. Speci ica ion Tes s in Econome ics. Econome ica, 46, 1251-
1271.
Heady, C., 2003. The E ec o Child Labo on Lea ning Achie emen . Wo ld
De elopmen 31, 385-398.
Jacoby, H., Skou ias, E., 1996. Risk, inancial ma ke s, and human capi al in a
de eloping coun y. Re iew o Economic S udies 64, 311–335.
Knodel, J., Wongsi h, M., 1991. Family size and child en’s educa ion in Thailand:
E idence om a na ional sample. Demog aphy 28, 119–131.
Liu, H. W., 1959. An analysis o Chinese clan ules: Con ucian heo ies in ac ion.
S an o d: S an o d Uni e si y P ess.
Lu, F, Ande son M., 2015. Pee E ec s in Mic oen i onmen s: he Bene i s o
Homogeneous Class oom G oups. Jou nal o Labo Economics 33, 91–122.
McDonald, J. F., Mo i , R., 1980. The uses o Tobi analysis. The Re iew o
Economics and S a is ics, 62, 318–321.
30
Moehling, C., 2004. Family s uc u e, school a endance, and child labo in Ame ican
sou h in 1900 and 1910. Explo a ions in Economic His o y 41, 73–100.
Mundlak, Y., 1978. On he pooling o ime se ies and c oss sec ion da a.
Econome ica, 46(1), 69-85.
Munshi, K., Rosenzweig, M., 2006. T adi ional Ins i u ions Mee he Mode n Wo ld:
Cas e, Gende , and Schooling Choice in a Globalizing Economy. The Ame ican
Economic Re iew, 96, 1225–1252
Nunn, N., Wan chekon, L., 2011. The Sla e T ade and he O igins o Mis us in
A ica. Ame ican Economic Re iew 101, 3221-3252.
O'Donnell, O., Rosa i, F.C., Doo slae , E.V., 2005. Heal h e ec s o child wo k:
E idence om u al Vie nam. Jou nal o Popula ion Economics 18, 437-467.
Os e , E., 2019. Unobse able selec ion and coe icien s abili y: Theo y and
e idence. Jou nal o Business & Economic S a is ics, 37, 187–204.
Peng, Y., 2004. Kinship Ne wo ks and En ep eneu s in China’s T ansi ional
Economy1. Ame ican Jou nal o Sociology 109, 1045-1074.
Pi , M. M., Rosenzweig, M. R., Hassan, N., 2012. Human Capi al In es men and
he Gende Di ision o Labo in a B awn-Based Economy. The Ame ican
Ranjan, P., 2001. C edi cons ain s and he phenomenon o child labo . Jou nal o
De elopmen Economics 64, 81-102.
Rosenbaum, PR., 2007. In e e ence be ween Uni s in Randomized Expe imen s.
Jou nal Ame ican S a is ical Associa ion 102, 191–200.
Tabellini, G., 2008. The Scope o Coope a ion: Values and Incen i es. Qua e ly
Jou nal o Economics 123, 905-950.
Tang, C., Zhao, L., Zhao, Z., 2018. Child Labo in China. China Economic Re iew
51, 149-166.
Tang, C., Zhao, L., Zhao, Z., 2020. Does F ee Educa ion Help Comba Child Labo ?
The E ec o a F ee Compulso y Educa ion Re o m in Ru al China. Jou nal o
Popula ion Economics 33, 601-631.
Tsai, L.L., 2007. Solida y G oups, In o mal Accoun abili y, and Local Public Goods
P o ision in Ru al China. Ame ican Poli ical Science Re iew 101, 355-372.
Ud y, C.R., 2003. Child Labo . Wo king Pape s 3, 79-79.
Voig lände , N., Vo h, H.J., 2012. Pe secu ion Pe pe ua ed: The Medie al O igins o
An i-Semi ic Violence in Nazi Ge many*. Qua e ly Jou nal o Economics 127,
1339-1392.
Wahba, J., 2006. The in luence o ma ke wages and pa en al his o y on child labou
and schooling in Egyp . Jou nal o Popula ion Economics 19, 823–852.
31
Wa son, J.L., 1982. Chinese Kinship Reconside ed: An h opological Pe spec i es on
His o ical Resea ch. China Qua e ly 92, 589-622.
Webe , M., 1981. Gene al Economic His o y. New B unswick, N.J.: T ansac ion
Books.
Xie, Y., 2012. China amily panel s udies (2010) Use 's manual. Ins i u e o Social
Science Su ey: Peking Uni e si y, Beijing, China.
Xu, Y., Yao, Y., 2015. In o mal Ins i u ions, Collec i e Ac ion, and Public
In es men in Ru al China. Ame ican Poli ical Science Re iew 109, 371-391.
Young HP., 2015. The e olu ion o social no ms. The Annual Re iew o Economics
7, 359–87.
Zhang, C., 2019. Family Suppo o Social Suppo ? The Role o Clan Cul u e.
Jou nal o Popula ion Economics 32, 529-549.
Zhang, C., 2020. Clans, En ep eneu ship, and De elopmen o he P i a e Sec o in
China. Jou nal o Compa a i e Economics 48, 100-123.
Zhao, L., Wang, F., Zhao, Z., 2021. T ade Libe aliza ion and Child Labo . China
Economic Re iew 65.
32
Figu e 1. Geog aphical Dis ibu ion o Clan Cul u e
No es: 1. This igu e shows he p opo ion o households holding genealogy books a
p o ince le el;
2. Au ho s’ calcula ion based on he amily ques ionnai e o CFPS 2010.
33
Figu e 2. Child Labo Engagemen o Boys, p- alue=0.033
No e: Es ima ed coe icien s om he pe mu a ion placebo es s. We andomly assign whe he o
no he e is a genealogy book o he child en’s households as placebo ea men s a us o he same
sample in he baseline analysis. These his og ams display he dis ibu ion o placebo ea men
e ec s om 4000 andom assignmen s. The dashed line shows he es ima ed ea men e ec om
he baseline analysis. The p alue o each pe mu a ion placebo es is he p opo ion o placebo
es ima es ha a e equal o o la ge in absolu e alue han he co esponding es ima e om he
baseline analysis.

34
Table 1. S eng h o Consanguini y Ties
Va iables
（1）
（2）
（3）
（4）
（5）
Households wi hou genealogies
Households wi h genealogies
Di e ence
Obs
Mean
Obs
Mean
Pa icipa e in ances o wo ship/ omb-sweeping
ac i i ies las yea (dummy)
11378
0.676
3322
0.766
-0.090***
How o en did he amily do he ollowing
con ac s wi h ela i es and iends las mon h
：
En e ainmen / dine oge he
11369
0.575
3319
0.723
-0.148***
Gi e ood o gi s
11374
0.381
3322
0.568
-0.187***
Help each o he
11371
0.384
3319
0.497
-0.114***
Pay a isi
11368
0.742
3320
0.888
-0.146***
Cha
11367
1.147
3321
1.287
-0.141***
No es: 1. F equencies o con ac s a e di ided in o 5 ca ego ies: 0, no con ac a all; 1, once a mon h; 2, 2-3 imes a mon h; 3, 2-3
imes a week; 4, almos e e y day.
2. Au ho s’ calcula ion based on he amily ques ionnai e o CFPS2010.
3. T- es o di e ence be ween households wi h a genealogy book and wi hou is epo ed in Column (5).
4. Resul s a e weigh ed using sample weigh s p o ided in he da a.5. *** p<0.01, ** p<0.05, * p<0.10.
35
Table 2. Summa y S a is ics
(1)
(2)
(3)
(4)
(5)
Va iables
Obs
Mean
S d. De .
Min
Max
Child labo (dummy)
3435
0.078
0.268
0
1
Genealogy (dummy)
3398
0.289
0.453
0
1
Age
3435
12.473
1.711
10
15
Han na ionali y (dummy)
3435
0.835
0.371
0
1
Boy (dummy)
3435
0.537
0.499
0
1
Ru al Hukou (dummy)
3405
0.772
0.419
0
1
In e p o incial mig an (dummy)
3426
0.014
0.116
0
1
Educa ion o a he (yea )
3356
7.015
4.288
0
19
Educa ion o a he (ca ego y)
Illi e acy
3382
0.171
0.376
0
1
P ima y School
3382
0.271
0.445
0
1
Junio High School
3382
0.374
0.484
0
1
Senio High School
3382
0.124
0.33
0
1
College and abo e
3382
0.06
0.238
0
1
Educa ion o mo he (yea )
3357
5.625
4.563
0
22
Educa ion o mo he (ca ego y)
Illi e acy
3386
0.3
0.458
0
1
P ima y School
3386
0.273
0.446
0
1
Junio High School
3386
0.298
0.458
0
1
Senio High School
3386
0.083
0.276
0
1
College and abo e
3386
0.045
0.208
0
1
Age o a he
3404
40.257
4.995
28
75
Age o mo he
3389
38.375
4.619
24
72
Household head is male (dummy)
3,426
0.758
0.428
0
1
Numbe o child en
3431
1.824
0.926
1
7
Numbe o adul s
3431
2.858
1.193
1
11
Household ne asse s (millions o CNY)
3238
0.168
0.470
-0.586
29.961
Non-ag icul u al ac i i ies by he amily (dummy)
3428
0.1
0.299
0
1
Numbe o ela i es isi ed las sp ing es i al
3410
5.781
6.458
0
100
A e age household ne asse s in he
illage/communi y (millions o CNY)
3428
0.170
0.224
-0.024
5.313
A e age educa ion yea o adul s in he
illage/communi y
3435
5.838
2.49
0.667
14.231
Popula ion in he illage/communi y
3378
4186.825
4663.566
170
51139
No es: 1. Au ho s’ calcula ion based on he amily ques ionnai e o CFPS 2010.
2. Resul s a e weigh ed using sample weigh s p o ided in he da a.
36
Table 3. Summa y S a is ics: by Ha e a Genealogy Book o No
Va iables
Genealogy=0
Genealogy=1
Di e ence
(1)
(2)
(3)
(4)
(5)
Obs
Mean
Obs
Mean
Child labo (dummy)
2501
0.091
897
0.046
-0.046***
Wo k hou s pe day
2501
0.618
897
0.302
-0.316***
Age
2501
12.494
897
12.433
-0.061
Han na ionali y (dummy)
2501
0.818
897
0.878
0.060***
Boy (dummy)
2501
0.532
897
0.553
0.021
Ru al Hukou (dummy)
2477
0.755
893
0.816
0.061***
Educa ion o a he (yea )
2439
6.839
881
7.409
0.570***
Educa ion o mo he (yea )
2451
5.539
872
5.870
0.331
Age o a he
2479
40.262
889
40.197
-0.065
Age o mo he
2470
38.427
883
38.249
-0.178
Household head is male (dummy)
2501
0.728
895
0.831
0.103***
Numbe o child en
2500
1.778
894
1.933
0.155***
Numbe o adul s
2500
2.852
894
2.864
0.012
Household ne asse s (millions o CNY)
2355
0.161
857
0.186
0.025
Non-ag icul u al ac i i ies by he amily
(dummy)
2501
0.101
897
0.098
-0.002
Numbe s o ela i es isi ed las sp ing es i al
2488
5.317
892
6.921
1.605***
A e age household ne asse s in he
illage/communi y (millions o CNY)
2497
0.172
894
0.156
-0.016**
A e age educa ion yea o adul s in he
illage/communi y
2501
5.855
897
5.793
-0.062
Log Popula ion in he illage/communi y
2451
7.904
891
7.875
-0.029
No es: 1. Au ho s’ calcula ion based on he amily ques ionnai e o CFPS2010.
2. T- es o di e ence be ween child en om amily wi h a genealogy book and wi hou is
epo ed in Column (5).
3. Resul s a e weigh ed using sample weigh s p o ided in he da a.
37
Table 4. Impac s o Clan Cul u e on Incidence o Child Labo
(1)
(2)
(3)
(4)
All
Pa ia chal and Pa ilineal
VARIABLES
All
Gi l
Boy
Genealogy*Boy
-0.039*
(0.021)
Genealogy (dummy)
-0.027**
-0.006
-0.016
-0.039**
(0.012)
(0.018)
(0.020)
(0.015)
Age
-0.089
-0.088
-0.134
-0.074
(0.059)
(0.059)
(0.093)
(0.076)
Age squa ed
0.003
0.003
0.005
0.003
(0.002)
(0.002)
(0.004)
(0.003)
Han na ionali y (dummy)
-0.035*
-0.035*
-0.066**
-0.001
(0.021)
(0.021)
(0.033)
(0.027)
Boy (dummy)
-0.011
0.000
(0.011)
(0.013)
Ru al Hukou (dummy)
-0.006
-0.005
0.002
0.005
(0.019)
(0.019)
(0.025)
(0.026)
Age o a he
-0.006
-0.006
0.053**
-0.028
(0.018)
(0.018)
(0.021)
(0.021)
Age o a he squa ed
0.000
0.000
-0.001**
0.000
(0.000)
(0.000)
(0.000)
(0.000)
Age o mo he
0.030**
0.029**
0.001
0.043***
(0.014)
(0.014)
(0.028)
(0.015)
Age o mo he squa ed
-0.000**
-0.000*
0.000
-0.000***
(0.000)
(0.000)
(0.000)
(0.000)
Household head is male (dummy)
0.012
0.012
0.020
0.013
(0.014)
(0.013)
(0.018)
(0.020)
Numbe o child en
0.003
0.002
0.004
-0.004
(0.010)
(0.010)
(0.014)
(0.013)
Numbe o adul s
-0.009
-0.009
-0.012
-0.004
(0.006)
(0.006)
(0.009)
(0.007)
Household ne asse s
0.003
0.004
-0.001
0.000
(0.007)
(0.007)
(0.024)
(0.007)
Non-ag icul u al ac i i ies by he amily (dummy)
-0.021
-0.022
0.000
-0.033*
(0.015)
(0.015)
(0.027)
(0.018)
Numbe o ela i es isi ed las sp ing es i al
-0.001
-0.001
0.002
-0.006***
(0.001)
(0.001)
(0.002)
(0.002)
A e age household ne asse s in he
illage/communi y
-0.034
-0.033
-0.111
0.014
(0.045)
(0.045)
(0.098)
(0.049)
A e age educa ion yea o adul s in he
illage/communi y
-0.014***
-0.014***
-0.002
-0.020***
(0.005)
(0.005)
(0.009)
(0.006)
Log popula ion in he illage/communi y
0.025**
0.025**
0.019
0.019
(0.010)
(0.010)
(0.016)
(0.013)
Cons an
-0.021
-0.022
-0.438
0.157
(0.376)
(0.376)
(0.638)
(0.464)
Obse a ions
3,023
3,023
1,509
1,514
R-squa ed
0.251
0.252
0.323
0.321
Coun y FE
Y
Y
Y
Y
Educa ion o Mo he (ca ego y)
Y
Y
Y
Y
Educa ion o Fa he (ca ego y)
Y
Y
Y
Y
No es: 1. The dependen a iable is child en’s labo o ce pa icipa ion.
2. Es ima es a e weigh ed using sample weigh s p o ided in he da a.
3. Robus s anda d e o s in pa en heses a e clus e ed a household le el.
4. *** p<0.01, ** p<0.05, * p<0.10.
44
Table 11. No h-Sou h Di e ence
(1)
(2)
(3)
(4)
(5)
(6)
VARIABLES
No h
No h&Gi l
No h&Boy
Sou h
Sou h&Gi l
Sou h&Boy
Genealogy (dummy)
0.007
0.029
-0.028
-0.045**
-0.031
-0.045**
(0.016)
(0.024)
(0.019)
(0.018)
(0.027)
(0.022)
Obse a ions
1,589
805
784
1,434
704
730
R-squa ed
0.172
0.280
0.235
0.299
0.367
0.378
Indi idual cha .
Y
Y
Y
Y
Y
Y
Household cha .
Y
Y
Y
Y
Y
Y
Village/Communi y cha .
Y
Y
Y
Y
Y
Y
Coun y FE.
Y
Y
Y
Y
Y
Y
No es: 1. The dependen a iable is child en’s labo o ce pa icipa ion.
2. Es ima es a e weigh ed using sample weigh s p o ided in he da a.
3. Robus s anda d e o s in pa en heses a e clus e ed a household le el.
4. *** p<0.01, ** p<0.05, * p<0.10.

45
Table 12. Mechanism: A Supplemen o Fo mal Ins i u ions
(1)
(2)
(3)
(4)
(5)
(6)
VARIABLES
U ban
U ban&Gi l
U ban&Boy
Ru al
Ru al&Gi l
Ru al&Boy
Panel A: by Hukou S a us
Genealogy (dummy)
-0.024
0.007
-0.019
-0.025*
-0.014
-0.039**
(0.030)
(0.040)
(0.033)
(0.014)
(0.021)
(0.019)
Obse a ions
612
307
305
2,411
1,202
1,209
R-squa ed
0.457
0.716
0.627
0.271
0.363
0.314
Panel B: by Residence
Genealogy (dummy)
-0.040
-0.049
-0.035
-0.027**
-0.009
-0.044**
(0.028)
(0.033)
(0.025)
(0.012)
(0.023)
(0.020)
Obse a ions
715
343
337
3,023
1,136
1,146
R-squa ed
0.398
0.656
0.516
0.251
0.353
0.315
No es: 1. The dependen a iable is child en’s labo o ce pa icipa ion.
2. Es ima es a e weigh ed using sample weigh s p o ided in he da a.
3. Robus s anda d e o s in pa en heses a e clus e ed a household le el.
4. *** p<0.01, ** p<0.05, * p<0.10.
46
Table 13. Mechanism: Impac s o Clan Cul u e by Household Socioeconomic S a us
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Ne asse <1/3
1/3<=Ne asse <2/3
Ne asse s>=2/3
VARIABLES
All
Gi l
Boy
All
Gi l
Boy
All
Gi l
Boy
Genealogy (dummy)
-0.084***
-0.053
-0.101***
0.004
-0.008
0.014
-0.003
-0.004
-0.024
(0.021)
(0.038)
(0.030)
(0.026)
(0.042)
(0.038)
(0.024)
(0.042)
(0.030)
Obse a ions
1,004
510
494
1,007
496
511
1,007
502
505
R-squa ed
0.354
0.515
0.426
0.358
0.485
0.448
0.468
0.624
0.619
Indi idual cha .
Y
Y
Y
Y
Y
Y
Y
Y
Y
Household cha .
Y
Y
Y
Y
Y
Y
Y
Y
Y
Village/Communi y cha .
Y
Y
Y
Y
Y
Y
Y
Y
Y
Coun y FE.
Y
Y
Y
Y
Y
Y
Y
Y
Y
No es: 1. The dependen a iable is child en’s labo o ce pa icipa ion.
2. Es ima es a e weigh ed using sample weigh s p o ided in he da a.
3. Robus s anda d e o s in pa en heses a e clus e ed a household le el.
4. *** p<0.01, ** p<0.05, * p<0.10.
47
Table 14. Mechanism: Risk-sha ing
(1)
(2)
(3)
VARIABLES
All
Gi l
Boy
Genealogy (dummy)
0.025
0.034
0.015
(0.024)
(0.039)
(0.030)
Disas e losses a p o ince le el
0.004***
0.004***
0.005***
(0.001)
(0.001)
(0.001)
Genealogy* Disas e losses a p o ince le el
-0.003***
-0.003
-0.003**
(0.001)
(0.002)
(0.001)
(0.001)
(0.001)
(0.001)
Obse a ions
3,023
1,509
1,514
R-squa ed
0.111
0.113
0.144
Indi idual cha .
Y
Y
Y
Household cha .
Y
Y
Y
Village/Communi y cha .
Y
Y
Y
P o incial cha .
Y
Y
Y
Region FE
Y
Y
Y
No es: 1. The dependen a iable is child en’s labo o ce pa icipa ion.
2. Es ima es a e weigh ed using sample weigh s p o ided in he da a.
3. Robus s anda d e o s in pa en heses a e clus e ed a household le el.
4. P o incial cha ac e is ics includes log o GDP and log o popula ion in 2009.
5. Six egions: No heas , No hwes , No h China, Sou hwes , Cen al Sou h, and Eas China.
6. *** p<0.01, ** p<0.05, * p<0.10.
48
Table 15. Mechanism: Easing C edi Cons ain s
(1)
(2)
(3)
(4)
(5)
(6)
Fa he
Mo he
VARIABLES
All
Gi l
Boy
All
Gi l
Boy
Panel A: Ge ing aids om ela i es
Genealogy
0.023
-0.033
0.080*
-0.020
-0.070
0.001
(0.035)
(0.050)
(0.048)
(0.033)
(0.046)
(0.045)
Obse a ions
2,245
1,106
1,139
2,530
1,264
1,266
R-squa ed
0.207
0.305
0.266
0.204
0.305
0.265
Panel B: Financial suppo
Genealogy
0.024
-0.042
0.079*
-0.054*
-0.089**
-0.038
(0.035)
(0.049)
(0.048)
(0.030)
(0.040)
(0.042)
Obse a ions
2,221
1,097
1,124
2,497
1,251
1,246
R-squa ed
0.209
0.314
0.273
0.219
0.317
0.258
Panel C: Aids on o he a ai s
Genealogy
-0.034
-0.014
-0.044
0.021
-0.001
0.036
(0.026)
(0.038)
(0.033)
(0.026)
(0.039)
(0.034)
Obse a ions
2,245
1,106
1,139
2,530
1,264
1,266
R-squa ed
0.158
0.211
0.242
0.162
0.231
0.234
No es: 1. The dependen a iable is whe he a he s/mo he s ecei e aids o he
ollowing a ai s: bo owing money; child en’s schooling; seeing a doc o ; job sea ching;
child en’s job sea ching.
2. Es ima es a e weigh ed using sample weigh s p o ided in he da a.
3. Robus s anda d e o s in pa en heses a e clus e ed a household le el.
4.*** p<0.01, ** p<0.05, * p<0.10.
49
Table 16. Mechanism: Social No ms
(1)
(2)
(3)
(4)
(5)
(6)
Log educa ional expendi u e (OLS)
S udy ime (Tobi )
VARIABLES
All
Gi l
Boy
All
Gi l
Boy
Genealogy (dummy)
0.222**
0.003
0.318**
0.286**
0.188
0.314*
(0.103)
(0.159)
(0.130)
(0.143)
(0.217)
(0.169)
Obse a ions
3,007
1,500
1,507
3,023
1,509
1,514
R-squa ed
0.367
0.437
0.404
Indi idual cha .
Y
Y
Y
Y
Y
Y
Household cha .
Y
Y
Y
Y
Y
Y
Village/Communi y cha .
Y
Y
Y
Y
Y
Y
Coun y FE.
Y
Y
Y
Y
Y
Y
Log pseudolikelihood
-182000000
-83800000
-94900000
Pseudo R-squa e
0.0797
0.0952
0.0952
No es: 1. The dependen a iable in Columns (1)-(3) is log household educa ional spending, and in Columns (4)-(6)
is hou s spending on s udy. Es ima es a e weigh ed using sample weigh s p o ided in he da a.
2. Robus s anda d e o s in pa en heses a e clus e ed a household le el.
3. *** p<0.01, ** p<0.05, * p<0.10.

TheUNU‐MERITWORKINGPape Se ies

2022-01 S uc u al ans o ma ionsandcumula i ecausa ion owa dsane olu iona y
mic o‐ ounda iono  heKaldo iang ow hmodelbyAnd éLo en z,TommasoCia li,
Ma iaSa onaandMa coValen e
2022-02 Es ima iono ap oduc ion unc ionwi hdomes icand o eigncapi als ockby
ThomasZieseme 
2022-03 Au oma ionand ela ed echnologies:Amappingo  henewknowledgebaseby
En icoSan a elli,JacopoS accioliandMa coVi a elli
2022-04 Theold‐agepensionhousehold eplacemen  a einBelgiumbyAlessioJ.G.B own
andAnne‐Lo eF aikin
2022-05 Globalisa ioninc eased us inno he nandwes e nEu opebe ween2002and
2018byLoesjeVe hoe enandJoRi zen
2022-06 Globalisa ionand inancialisa ionin heNe he lands,1995–2020byJoanMuysken
andHuubMeije s
2022-07 Impo pene a ionandmanu ac u ingemploymen :E idence omA icaby
SolomonOwusu,GideonNdubuisiandEmmanuelB.Mensah
2022-08 Ad anceddigi al echnologiesandindus ial esiliencedu ing heCOVID‐19
pandemic:A i m‐le elpe spec i ebyElisaCalzaAlejand oLa opaandLigiaZaga o
2022-09 The eckoningo sexual iolenceandco up ion:Agende eds udyo sex o ionin
mig a ion oSou hA icabyAshleighBicke Caa en,Loes anHeug enandO un
Me kle
2022-10 Thep oduc i e oleo socialpolicybyOma Rod íguezTo es
2022-11 Somenew iewsonp oduc spaceand ela eddi e si ica ionbyÖnde Nomale 
andBa Ve spagen
2022-12 Themul idimensionalimpac so  heCondi ionalCashT ans e p og amJun osin
Pe ubyRica doMo elandLizGi ón
2022-13 Semi‐endogenousg ow hinanon‐Wal asianDSEM o B azil:Es ima ionand
simula iono changesin o eignincome,humancapi al,R&D,and e mso  ade
byThomasH.W.Zieseme 
2022-14 Rou ine‐biased echnologicalchangeandemployeeou comesa e masslayo s:
E idence omB azilbyAn onioMa ins‐Ne o,Xa ie Ci e aandAlexCoad
2022-15 Thecanonicalco ela ioncomplexi yme hodbyÖnde Nomale &Ba Ve spagen
2022-16 Canonicalco ela ioncomplexi yo Eu opean egionsbyÖnde Nomale &Ba 
Ve spagen
2022-17 Quan ile e u nand ola ili yconnec ednessamongNon‐FungibleTokens(NFTs)
and(un)con en ionalasse sbyCh is ianU om,GideonNdubuisiandKhaled
Guesmi
2022-18 HowdoFi msInno a einLa inAme ica?Iden i ica iono Inno a ionS a egiesand
Thei MainAdop ionDe e minan sbyFe nandoVa gas
2022-19 Remi ancedependence,suppo  o  axa ionandquali yo publicse icesinA ica
byMa yKon eandGideonNdubuisi
2022-20 Ha monizedLa inAme icaninno a ionSu eysDa abase(LAIS):Fi m‐le el
mic oda a o  hes udyo inno a ionbyFe nandoVa gas,Cha lo eGuilla d,
MónicaSalaza andGus a oA.C espi
2022-21 Au oma ionexposu eandimplica ionsinad ancedandde elopingcoun iesac oss
gende ,age,andskillsbyHube Nii‐Aponsah
2022-22 Sex o ioninaccess oWASHse icesinselec ed egionso BangladeshbyO un
Me kle,Um bekAllakulo andDebo aGonzalez
2022-23 Complexi y esea chineconomics:pas ,p esen and u u ebyÖnde Nomale &
Ba Ve spagen
2022-24 Technologyadop ion,inno a ionpolicyandca ching‐upbyJuanR.Pe illaand
ThomasH.W.Zieseme 
2022-25 Exogenousshocksandp oac i e esiliencein heEU:Thecaseo  heReco e yand
ResilienceFacili ybyAn honyBa zokas,Rena oGiaconandCo adoMacchia elli
2022-26 Achance o op imism:Enginee ing heb eakaway om hedownwa dspi alin
us andsocialcohesiono keeping he ish omdisappea ingbyJoRi zenand
Eleono aNillesen
2022-27 Dynamicdependencebe weencleanin es men sandeconomicpolicyunce ain y
byCh is ianU om,HelaMzoughi,GideonNdubuisiandKhaledGuesmi
2022-28 Pee ne wo ksandmalleabili yo educa ionalaspi a ionsbyMichelleGonzalez
Amado ,RobinCowanandEleono aNillesen
2022-29 Linking heBOPCg ow hmodelwi h o eigndeb dynamics o hegoodsandlabou 
ma ke s:ABOP‐IXSM‐OkunmodelbyThomasH.W.Zieseme 
2022-30 Coun ies’ esea chp io i iesin ela ion o heSus ainableDe elopmen Goalsby
HugoCon a ia,TommasoCia liandEdNoyons
2022-31 Mul i askingbyAnzelikaZaice a‐Razzolini
2022-32 In o malins i u ionmee schildde elopmen :Clancul u eandchildlabo inChina
byCanTangandZhongZhao