Washing machine ownership and girls' school attendance: a cross-sectional analysis of adolescents in 19 middle-income countries

Ka lsson, Oma ; De Ne e, Jan-Wal e
A icle — Published Ve sion
Washing machine owne ship and gi ls' school a endance:
a c oss-sec ional analysis o adolescen s in 19 middle-
income coun ies
The Jou nal o Economic Inequali y
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: Ka lsson, Oma ; De Ne e, Jan-Wal e (2024) : Washing machine owne ship and
gi ls' school a endance: a c oss-sec ional analysis o adolescen s in 19 middle-income coun ies,
The Jou nal o Economic Inequali y, ISSN 1573-8701, Sp inge US, New Yo k, NY, Vol. 22, Iss. 3, pp.
735-755,
h ps://doi.o g/10.1007/s10888-023-09612-7
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/315324
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
Vol.:(0123456789)
h ps://doi.o g/10.1007/s10888-023-09612-7
1 3
Washing machine owne ship andgi ls’ school a endance:
ac oss‑sec ional analysis o adolescen s in19 middle‑income
coun ies
Oma Ka lsson1,2 · Jan‑Wal e DeNe e3
Recei ed: 23 Augus 2023 / Accep ed: 8 Decembe 2023
© The Au ho (s) 2024
Abs ac
Excessi e wo k among adolescen s may comp omise educa ional de elopmen . Wi h-
ou home appliances, household wo k can ake o e 50h a week and an addi ional 30h
when an in an is p esen . School-aged gi ls a e o en asked wi h doing laund y, which is
ime-consuming and in lexible wi hou a washing machine. We de e mined he associa-
ion be ween washing machine owne ship and school a endance among adolescen s ages
10–19yea s in 19 middle-income coun ies be ween 2000 and 2021 (N = 1,622,514). We
con olled o socioeconomic and demog aphic cha ac e is ics, all neighbo hood-le el ac-
o s, and examined di e ences by sex, age, household weal h, and pe iod. No ela ionship
be ween washing machine owne ship and school a endance was ound in mos coun ies:
Howe e , he e was a subs an ial associa ion o gi ls in Tü kiye and a small o mode a e
associa ion o gi ls in Egyp and Albania. In Tü kiye, o example, gi ls li ing in house-
holds wi h a washing machine had 28% (95% CI 19, 37) g ea e school a endance com-
pa ed o gi ls li ing in households which did no . No associa ion was obse ed o boys.
The esul s sugges ha household owne ship o a washing machine does gene ally no
imp o e school a endance among gi ls, excep possibly in speci ic con ex s.
Keywo ds School a endance· Household wo k· Washing machine· Middle-income
coun ies
* Oma Ka lsson
oma [email p o ec ed]
Jan-Wal e De Ne e
janwal e .dene [email p o ec ed]
1 Visi ing Resea ch Fellow, Duke Uni e si y Popula ion Resea ch Ins i u e, Duke Uni e si y, 140
Science D , Du ham, NC27710, USA
2 Cen e o Economic Demog aphy, School o Economics andManagemen , Lund Uni e si y, P.O.
Box7083, 22007Lund, Sweden
3 Heidelbe g Ins i u e o Global Heal h, Medical Facul y andUni e si y Hospi al, Uni e si y
o Heidelbe g, Im Neuenheime Feld 130.3, 69120Heidelbe g, Ge many
The Jou nal o Economic Inequali y (2024) 22:735–755
/ Published online: 27 Janua y 2024
O.Ka lsson, J.-W.De Ne e
1 3
1 In oduc ion
Schooling is widely held as a co ne s one o economic as well as human de elopmen (De
Ne e e al. 2020). Fo indi iduals, acqui ing educa ion imp o es hei own and hei chil-
d en’s income, heal h, and well-being (Ka lsson e al. 2018; De Ne e e al. 2015). An edu-
ca ed popula ion helps inc ease en ep eneu ial ac i i y, economic g ow h, and ax e enue
o inance essen ial public in as uc u e. I also p o ides skills o illing he inc easingly
complex oles coun ies need o s ay compe i i e and a ain he highes li ing s anda ds
o hei ci izens. Gi ls’ educa ion has his o ically lagged behind ha o boys and con in-
ues o do so in many coun ies (UNESCO 2022), while emale educa ion is belie ed o
hold pa icula po en ial, o example, o imp o ing child heal h (Caldwell 1979). The e-
o e, nume ous de elopmen goals cen e on educa ion wi h a special ocus on gi ls (WHO
2015).
UNESCO (2022) es ima ed ha 89% o he ele an age g oup was en olled in p ima y
educa ion in 2018 globally. Howe e , he en olmen a e ell o 66% o seconda y school.
Many ac o s con ibu e o he all in en ollmen a e p ima y school, such as poo school
quali y, long a el imes om he household o school, and lack o inances o school ees.
Po e y is an unde lying cause (Jabba ian e al. 2022). "Child labo "—de ined as "wo k
ha dep i es child en o hei childhood, hei po en ial and hei digni y, and ha is ha m-
ul o physical and men al de elopmen " and "in e e es wi h hei schooling"—is ano he
common consequence o po e y ha nega i ely a ec s schooling in many coun ies (ILO
2022b; Pu nick and Bo ns ein 2015). School-aged gi ls a e mo e o en asked wi h wo k
inside he home—such as cooking, cleaning, and laund y—while boys wo k ou side he
home o in amily p oduc ion (i.e., on a amily a m), mi o ing o he gende ed wo k.
Household appliances can po en ially educe he wo kload om household wo k
(Cowan 1983; Moky 2000) and he e o e educe he need o employing child en wi hin
he household, enabling hem o a end school. Fo example, in a s udy om China,
12–18-yea -old gi ls spen 102min less on household wo k pe week and we e 17% mo e
likely o a end school when li ing in households wi h a washing machine, while he e ec
was much less p onounced among boys (Ke 2019). In India, owne ship o ime-sa ing
household appliances, such as a e ige a o , inc eased school en ollmen and dec eased
employmen a es among adolescen s ages 12–18yea s (Bha ga a and Ke 2021). In addi-
ion o educing he ime spen on household cho es, washing machines can also inc ease
school a endance by emo ing mic oo ganisms and he eby imp o ing child en’s physical
heal h – e.g., h ough washing clo hes a 130 deg ees Fah enhei o highe , pa icula ly in
con ex s whe e in ec ious diseases a e common among school-age child en (Ka lsson e al.
2020; Shi e al. 2022).
The impo ance o household appliances o gi ls’ school a endance is likely o be
highly con ex -dependen and a y acco ding o gende no ms, school access, and o he
ac o s, such as school quali y, e ili y, and he need o labo wi hin he home. Howe e ,
o ou knowledge, no s udy has comp ehensi ely examined he quan i a i e ela ionship
be ween owne ship o a washing machine and school a endance ac oss a wide ange o
con ex s. Fo example, no simila s udies ha e been iden i ied om he A ican egion,
whe e child en ela i ely equen ly engage in domes ic wo k a an ea ly age and school
a endance emains pe sis en ly low (G ae z e al. 2020). To begin o add ess his gap in
he li e a u e, his s udy he e o e examined he associa ion be ween owne ship o a wash-
ing machine and school a endance using na ionally ep esen a i e da a on 1.6 million
adolescen s li ing in 19 middle-income coun ies be ween 2000 and 2021. We s udied
736
Washing machine owne ship and gi ls’ school a endance
1 3
he e ogenei y in he associa ion ac oss sex, age, household weal h, and yea and adjus ed
o socioeconomic and demog aphic cha ac e is ics as well as all obse ed and unobse ed
ac o s a he le el o he adolescen ’s neighbo hood o illage.
2 Da a andme hods
2.1 Da a
The empi ical analyses p esen ed in his pape we e based on indi idual-le el da a om
he na ionally ep esen a i e Demog aphic and Heal h Su eys (DHS), which a e con-
duc ed egula ly in nume ous low- and middle-income coun ies. S anda dized ques ion-
nai es and measu es we e used so ha he su eys a e compa able be ween coun ies and
su ey yea s. The sampling was based on s a i ied mul i-s age sampling. S a i ica ion was
usually done by geog aphic o adminis a i e sub-na ional egions c ossed wi h he ype
o esidence (u ban o u al). P ima y sampling uni s, usually neighbo hoods o illages
(o en he enume a ion a eas o he mos ecen census), we e sampled wi hin each s a um
based on a p obabili y p opo ional o popula ion size. Households we e sampled om he
p ima y sampling uni s. Se e al ques ionnai es a e adminis e ed in each household. This
s udy only used da a om he household schedule, which collec s basic in o ma ion abou
each household membe —such as age, sex, and educa ion—and he household ques ion-
nai e, which collec s basic in o ma ion abou he household, such as owne ship o ce ain
asse s (e.g., washing machine, e ige a o , ele ision) and ameni ies (e.g., elec ici y, ype
o oile ).
The DHS esponse a es ha e been e y high, ypically exceeding 90% (Co si e  al.
2012). Non- esponses a e no eplaced, and smalle popula ion segmen s a e o e sampled
o acqui e adequa e obse a ions o analysis. Sampling weigh s—calcula ed as he in e se
p obabili y o being included in he su ey—a e p o ided o adjus o o e sampling and
non- esponses and o imp o e p ecision (Aliaga and Ren 2006). Addi ional in o ma ion on
he DHS su eys is a ailable elsewhe e (Aliaga and Ren 2006; Co si e al. 2012), as well as
DHS inal coun y epo s (DHS 2022).
2.2 S udy popula ion andinclusion c i e ion
Following he WHO (2022) de ini ion o adolescen s, we included 10–19-yea -olds. In o -
ma ion on owne ship o washing machines is no a pa o he co e ques ionnai e in he
DHS bu is some imes added in he su eys o mee local condi ions and needs (e.g., wash-
ing machine owne ship was p esumably added acco ding o coun y income le el o ob ain
app op ia e da a o cons uc ing an asse index). All su eys including his in o ma ion
we e conside ed o inclusion. The analysis was es ic ed o coun ies whe e a leas 5%
and a mos 95% o he analy ical sample a ended school o had a washing machine o
ensu e su icien a ia ion in he ou come and exposu e a iables.
Only usual esiden s o in e iewed households we e conside ed. The comple e sam-
ple consis ed o 1,622,514 adolescen s in 38 su eys om 19 coun ies: Albania, A me-
nia, Aze baijan, Colombia, Egyp , Gabon, Gua emala, Guyana, India, Indonesia, Ky gyz
Republic, Moldo a, Mo occo, Pakis an, Pe u, Philippines, Sou h A ica, Tajikis an, and
Tü kiye (Supplemen a y Table S1 and S2). These coun ies we e all classi ied by he
Wo ld Bank as middle-income a o a ound he ime when su eyed. A ew coun ies we e
737
O.Ka lsson, J.-W.De Ne e
1 3
s ill classi ied as low-income a he ime o su ey—speci ically, Pakis an in he ea lies o
h ee su eys, conduc ed in 2006 (became a middle-income coun y in 2008); Tajikis an in
he ea lie o wo su eys, conduc ed in 2012 (became a middle-income coun y in 2014);
he Ky gyz Republic, which was su eyed only in 2012 (became a middle-income coun y
in 2013); and Moldo a, which was only su eyed in 2005 (became a middle-income coun-
y in 2006). The numbe o su eys o each coun y anged om one o i e su eys. The
ea lies su ey was conduc ed in 2000, and he mos ecen su ey was in 2019–21. Mos
coun ies had mo e han one su ey, wi h Egyp ha ing he mos su eys ( i e).
We excluded obse a ions wi h missing alues o school a endance (n = 1,213), wash-
ing machine (he ea e also called washe ) owne ship (n = 538), co a ia es (n = 6,222),
and, las ly, neighbo hoods wi h only a single alid obse a ion (n = 758) since a leas
wo obse a ions a e needed wi hin a neighbo hood when compa ing neighbo s (see Sec-
ion2.6 o mo e in o ma ion on his c i e ion). The emaining sample o analysis com-
p ised 1,614,264 adolescen s 10–19 yea s old (Supplemen a y Table S1). The small-
es analy ical sample was om Moldo a (n = 5,457 om a single su ey om 2010) and
he la ges was om India (n = 1,075,968 om wo su eys conduc ed in 2015–16 and
2019–21).
2.3 Exposu e
The main independen a iable o in e es was a bina y indica o o whe he he house-
hold whe e he adolescen esided owned a washing machine. Mos su eys did no clea ly
dis inguish whe he he ques ion e e ed o au oma ic o manual washing machines. How-
e e , he In eg a ed Public Use Mic oda a Se ies (IPUMS) (2022) indica es ha hese a i-
ables e e o an au oma ic washing machine. Fu he , su ey epo s sugges ha hese
da a e e o au oma ic washing machines, calling hem "elec ical appliances," "mode n
con eniences," "appliances," and "du able goods" in hei inal epo s (DHS 2022). As a
sensi i i y analysis, desc ibed below, we also es ic he analy ical sample o adolescen s
li ing in households wi h access o elec ici y since elec ici y access is equi ed o au o-
ma ic washing machines (Supplemen a y Fig.S16 and TableS24).
We ocused on washing machines a he han o he ime-sa ing home appliances o
he ollowing eason: doing laund y is a pa icula ly ime-consuming and in lexible wo k,
commonly done by adolescen gi ls, which o en needs o be done by hand a e e ching
wa e o a eling o a wa e sou ce in he mo nings (du ing school hou s) o be d y and
eady by nigh ime (Assaad e al. 2010). Da a on he numbe o hou s spen doing domes-
ic and economic ac i i ies in he pas week was no consis en ly a ailable ac oss su eys.
We we e, he e o e, unable o s udy he associa ion be ween domes ic and economic ac i i-
ies o adolescen s, such as doing laund y o a ming, and school a endance.
2.4 Ou come
The main ou come s udied was a bina y indica o o whe he an adolescen 10–19yea s
old a ended school du ing he school yea in which he su ey ook place, cons uc ed
om he household schedule, which collec s basic in o ma ion on all household mem-
be s. The a iable was c ea ed om he ques ion: "Did [NAME] a end school o any ea ly
childhood educa ion p og am a any ime du ing he [YEAR] school yea ?" In addi ion, he
o al numbe o school yea s comple ed in he su ey yea was s udied as an ou come (see
Sec ion2.8 below).
738

Washing machine owne ship and gi ls’ school a endance
1 3
2.5 Con ol a iables
Con ol a iables we e added o accoun o demog aphic, socioeconomic, and spa ial di e -
ences which may con ound he ela ionship be ween washe owne ship and school a endance,
including idge owne ship, TV owne ship, ha ing a lush oile , a weal h index z-sco e (lin-
ea and squa ed), numbe o household membe s (linea and squa ed), numbe o household
membe s unde he age o i e (linea and squa ed), age o he adolescen (linea and squa ed),
highes educa ion le el o an adul male in he household, and highes educa ion le el o an
adul emale in he household, as well as all neighbo hood-le el ac o s. The highes educa-
ion le els o adul males and emales in he household only e e ed o adul s abo e age 25
and we e ca ego ized as 1) no educa ion, 2) p ima y, 3) seconda y, 4) highe , and 5) no adul
male/ emale in he household. The household weal h index—cons uc ed using p incipal com-
ponen analysis o a ious asse s and ameni ies—was p o ided wi h he DHS da ase s. The
weal h index ac o sco es we e con e ed in o su ey-speci ic z-sco es by sub ac ing an o e -
all (i.e., including all households in a su ey) weigh ed mean and di iding by he s anda d
de ia ion.
The weal h index was cons uc ed including washing machine owne ship and i s inclusion
as an independen a iable migh he e o e plausibly con ol away some o he e ec o wash-
ing machine owne ship on school a endance. Washing machine owne ship was, howe e ,
only one o many a iables used o cons uc he household weal h index, and indi idual i ems
gene ally do no con ibu e much o he index (Howe e al. 2012). Ne e heless, we add essed
his po en ial conce n in he Sec ion2.8 whe e we excluded weal h index om ou models.
Fu he , con olling o ha ing a lush oile , idge, and TV a home could be p oblem-
a ic due o mul icollinea i y o bad con ol p oblems—which would also obscu e an e ec o
washing machine owne ship on school a endance—since a la ge majo i y o hose who own
a washing machine also own a TV and a idge, and mos also ha e a lush oile (Supplemen-
a y TableS3). Howe e , calcula ing he a iance in la ion ac o o he main e ms in ou
main model does no indica e se ious mul icollinea i y (Supplemen a y TableS4). This issue
was also add essed wi h a sensi i i y analysis whe e hese co a ia es we e excluded om he
models (shown in he Sec ion2.8).
2.6 Analysis
Ou analysis p oceeded in wo s eps. Fi s , desc ip i e s a is ics show he p opo ion o ado-
lescen s a ending school (by sex) and li ing in households wi h washing machines. We also
show how hose p opo ions and owne ship o o he e e yday household i ems ha e e ol ed
o e ime, in he Supplemen .
Second, modi ied Poisson eg ession models (i.e., wi h sandwich s anda d e o s) we e
used o es ima e he associa ion be ween washe owne ship and school a endance (Zou
2004; Zou and Donne 2013). All models es ima ed school a endance (
y
) o an adolescen
(
i
) in neighbo hood (
n
) in e ms o being emale (
), washe owne ship, (
x1
) and in e ac ion
be ween being emale and washe owne ship (
×x1
), as well as con ol a iables (
x2…xJ
),
all also in e ac ed wi h being emale (
×xj
).
(1)
Poisson
(yin =1)=𝛼+𝛽 in +𝜎 n+
J
∑
j
=
1
[𝜌j( ×xj)in +𝜋jxj,in +𝛾j( ×xj)n+𝜏jxj,n
]
739
O.Ka lsson, J.-W.De Ne e
1 3
The model was adjus ed o neighbo hood-le el ac o s by adding a (weigh ed) neigh-
bo hood-le el mean o all independen a iables including in e ac ion e ms (
×x
,
x
, and
) as independen a iables (Schunck and Pe ales 2017; Woold idge 2019). The neigh-
bo hoods (o illages in u al a eas) we e indica ed by he p ima y sampling uni s. No e
ha we e EQ1 es ima ed using he same a iables bu a linea eg ession ins ead o Pois-
son eg ession, all pa ame e s below he neighbo hood le el would be iden ical i dummy
coded a iables o neighbo hoods we e added as independen a iables ins ead o he
neighbo hood le el means o all he independen a iables (al hough he o me app oach
migh no be compu a ionally easible due o he high numbe o neighbo hoods). How-
e e , since hese a e es ima ed using Poisson models, he pa ame e s di e be ween he wo
app oaches, mos likely only sligh ly, bu he in e p e a ion emains he same.
The exponen o he coe icien s o washe owne ship (
exp(𝜋1)
) gi es he a e a ios
o males, showing he ela i e di e ence in school a endance o males wi h a washe a
home compa ed o males wi hou (i.e., school a endance o hose wi h a washe di ided
by ha o hose wi hou a washe ). A a e a io o one means an equal a e. A a e a io
abo e one means g ea e school a endance o hose wi h a washe a home (e.g., a a e
a io o 1.2 means 1.2 imes he a e o 20% g ea e a e). A a e a io below one means
a lowe a e o hose wi h a washe a home (e.g., a a e a io o 0.4 means 0.4 imes
he a e o a 60% lowe a e). The exponen ia ed coe icien o he in e ac ion be ween
being emale and ha ing a washe (
exp(𝜌1)
) shows he ela i e di e ence in a e a io
o ha ing a washe o emales compa ed o males. Mul iplying he exponen ia ed coe -
icien o washe owne ship by he exponen ia ed in e ac ion e m (
exp[𝜌1]×exp[𝜋1]
o
exp[
𝜌
1
+𝜋
1]
) gi es he a e a io o emales (done using pos -es ima ion ollowing Eq.1).
Excep o he e m o being a emale, all independen a iables, including he neigh-
bo hood-le el a e ages, we e cen e ed a ound a (weigh ed) mean. The exponen ia ed con-
s an (
exp(𝛼)
), he e o e, shows school a endance o males wi h he mean on all o he
independen a iables. Mul iplying he exponen ia ed cons an by he exponen ia ed coe -
icien o emales (
exp(𝛼)×exp(𝛽)
o
exp(𝛼+𝛽)
) gi es he school a endance o emales
(done using pos -es ima ion ollowing Eq.1).
Es ima es we e weigh ed using sampling weigh s escaled o sum up o one in each su -
ey. Each su ey, he e o e, con ibu ed equally o he coun y es ima es a he han being
a ec ed by di e ences in sample sizes be ween su eys o changes in he unde lying popu-
la ion size. Weigh s o pooled es ima es we e u he escaled such ha each coun y con-
ibu ed equally o hese es ima es. Robus s anda d e o s used o cons uc 95% con i-
dence in e als we e adjus ed o clus e ing a he le el o p ima y sampling uni s.
2.7 S a i ica ion
Resul s we e u he s a i ied by age g oups (10–13yea s, 14–16yea s, and 17–19yea s)
and su ey yea ca ego ies. The s a i ica ion ac oss su ey yea s was only done o coun-
ies wi h mo e han one su ey. Since he yea o he su ey and leng h o ime be ween
su eys a ied, he pe iodiza ion was no he same in all coun ies: The su eys ell in h ee
pe iods: 2000–08, 2008–2016, and 2013–2021 (see Supplemen a y Table S5). A mos
h ee pe iods we e cons uc ed o each coun y, e en o coun ies wi h mo e han h ee
su eys. Finally, esul s we e s a i ied ac oss e ciles o he household weal h index (i.e.,
di iding he inal analy ical sample o each coun y in o h ee equally sized g oups acco d-
ing o he household weal h index).
740
Washing machine owne ship and gi ls’ school a endance
1 3
2.8 Supplemen a y andsensi i i y analyses
Fi s , we compa ed he ole o household owne ship o a washing machine wi h o he
household asse s and u ili ies wi h a ying ele ance o adolescen s’ ime-use and
school a endance. To do so, we conside ed he ela ionship be ween school a end-
ance and he p esence in he household o 1) a e ige a o (Supplemen a y Fig.S3 and
TableS11), 2) piped d inking wa e (Supplemen a y Fig.S4 and TableS12), 3) elec-
ici y (Supplemen a y Fig. S5 and Table S13), 4) a ca (Supplemen a y Fig.S6 and
TableS14), 5) a mo o bike (Supplemen a y Fig.S7 and TableS15), 6) clean cooking
uel (Supplemen a y Fig.S8 and TableS16), 7) a lush oile (Supplemen a y Fig.S9
and Table S17), and, las ly, 8) a TV (Supplemen a y Fig.S10 and Table S18). Mos
o hese i ems could ha e an impac on school a endance h ough simila pa hways as
washe owne ship, ha is by educing he bu den om household wo k (Supplemen a y
Appendix S1). Fu he , a po en ial impac may di e be ween boys and gi ls i he ype
o wo k ha is educed is gende ed. Howe e , owne ship o a TV seems unlikely o
a ec school a endance di e en ially be ween gi ls and boys—while s ill being linked
o household weal h—and can se e as a placebo es .
Two, es ima es we e ob ained using logi eg ession models (Supplemen a y Fig.S11
and TableS19) and linea eg ession models (Supplemen a y Fig.S12 and TableS20)
ins ead o Poisson eg ession models. Th ee, con ol a iables o he p esence o a lush
oile , idge, and TV in he household (Supplemen a y Fig.S13 and TableS21) and
he con ol a iable o household weal h index z-sco e (Supplemen a y Fig.S14 and
TableS22) we e excluded om he models. Fou , he sample was es ic ed o descend-
an s o he household head since o he s could po en ially include spouses o s a (Sup-
plemen a y Fig.S15 and TableS23). Fi e, we es ic ed he sample o households wi h
access o elec ici y (Supplemen a y Fig. S16 and Table S24). Six, he esul s we e
s a i ied acco ding o he o e all sha e o adolescen s a ending school (Supplemen a y
Fig.S17 and TableS25). Se en, he associa ion o washe owne ship wi h he yea s o
educa ion he adolescen had comple ed in he yea o he su ey was es ima ed, o e all
(Supplemen a y Fig.S18 and TableS26) and s a i ied ac oss su ey yea s (Supplemen-
a y Fig.S19 and TableS27).
Eigh , he esul s we e s a i ied acco ding o he ex en o emale disad an age in
school a endance a he egional le el (Supplemen a y Fig.S20 and TableS28). Nine,
esul s we e ob ained om a model adjus ing o household le el ac o s, es ima ing
he di e ence in he school a endance o boys and gi ls wi hin he same households,
in households wi h and wi hou a washe (Supplemen a y Fig.S21 and TableS29). The
models included baseline e ms only o being emale and age (since hese we e he only
independen a iables ha a ied wi hin households), as well as in e ac ion e ms o
emale wi h age and all o he independen a iables.
See Supplemen a y Appendix S1 o mo e de ails and esul s om supplemen a y
and sensi i i y analyses.
741
O.Ka lsson, J.-W.De Ne e
1 3
Table 1 Desc ip i e s a is ics
School a endance Washe owne ship Su ey yea s
Bo h Male Female Di e ence Bo h
Albania .836 .84 .833 − .0074 .853 2008–09, 2017–18
(N = 13,749) [.825, .847] [.827, .853] [.818, .848] [− .0246, .00979] [.835, .871]
A menia .85 .844 .856 .0114 .845 2005, 2010, 2015–16
(N = 10,925) [.84, .86] [.83, .859] [.842, .869] [− .00699, .0298] [.83, .859]
Aze baijan .871 .915 .827 − .0879*** .207 2006
(N = 6,203) [.857, .885] [.902, .929] [.807, .848] [− .11, − .0658] [.178, .235]
Colombia .799 .791 .808 .0165*** .431 2004–05, 2015–16
(N = 61,245) [.793, .805] [.783, .799] [.8, .815] [.00696, .026] [.416, .446]
Egyp .763 .788 .738 − .0496*** .809 2000, 2003, 2005, 2008, 2014
(N = 102,079) [.758, .769] [.782, .793] [.731, .745] [− .0563, − .0428] [.802, .815]
Gabon .902 .898 .905 .00617 .0722 2012
(N = 8,462) [.889, .914] [.882, .915] [.89, .919] [− .0134, .0257] [.0508, .0936]
Gua emala .684 .722 .646 − .0766*** .113 2014–15
(N = 24,453) [.671, .696] [.708, .736] [.63, .661] [− .0921, − .061] [.099, .127]
Guyana .771 .77 .773 .00341 .179 2005, 2009
(N = 6,801) [.756, .787] [.749, .791] [.753, .793] [− .0218, .0286] [.158, .2]
India .781 .797 .765 − .0315*** .13 2015–16, 2019–21
(N = 1,075,968) [.78, .783] [.795, .799] [.763, .767] [− .0338, − .0292] [.128, .133]
Indonesia .822 .813 .832 .0194*** .35 2017
(N = 35,687) [.816, .829] [.804, .822] [.824, .841] [.00795, .0308] [.335, .365]
Ky gyz Republic .91 .906 .913 .00738 .638 2012
(N = 6,551) [.897, .923] [.891, .92] [.897, .93] [− .0104, .0252] [.606, .671]
Moldo a .875 .85 .9 .0496*** .653 2005
(N = 5,457) [.863, .887] [.833, .867] [.886, .913] [.03, .0692] [.627, .679]
742
Washing machine owne ship and gi ls’ school a endance
1 3
95% CI 0.94, 1.35), bu he a e a io was no s a is ically signi ican . In Egyp , ha ing
a washe a home was associa ed wi h 6% (RR 1.06; 95% CI 1.04, 1.08) g ea e school
a endance among 10–12-yea -old gi ls and 10% (RR 1.10; 95% CI 1.06, 1.14) g ea e
among 13–15-yea -old gi ls, while he e was no di e ence o 16–19-yea -olds (RR 1, 95%
CI 0.96, 1.03). In A menia, ha ing a washe was associa ed wi h a 15% (RR 1.15; 95% CI
Fig. 4 Resul s om Poisson eg ession models o school a endance on washe owne ship: s a i ied by
age. No es: Coun ies we e o de ed acco ding o he ela i e di e ence in a e a io be ween boys and gi ls
om Fig.1. 95% con idence in e als a e shown. Es ima es we e weigh ed using sampling weigh s escaled
o sum up o one in each su ey. Pooled models we e u he escaled such ha each coun y con ibu ed
equally o he es ima es. All models included a baseline e m o being emale and baseline e ms and in e -
ac ions wi h being emale o washe owne ship, idge owne ship, TV owne ship, ha ing lush oile , a
weal h index z-sco e, numbe o household membe s, numbe o household membe s unde age i e, age,
highes educa ion le el o a male in household, and highes educa ion le el o a emale in household, as
well as adjus ing o neighbo hood. Uppe con idence limi s we e omi ed o es ima es abo e one and
lowe con idence limi s we e omi ed o es ima es below one, o imp o ed eadabili y. See Supplemen a y
TableS10 o abula ed es ima es
749

O.Ka lsson, J.-W.De Ne e
1 3
1.004, 1.31) g ea e school a endance o 16–19-yea -old gi ls, while he di e ence was
sligh (RR 1.03; 95% CI 0.97, 1.09 a ages 13–15) o non-exis en and no s a is ically sig-
ni ican a o he ages (RR 1; 95% CI 0.99, 1.02 a 10–12yea s).
3.6 Resul s omsupplemen a y andsensi i i y analyses
Ou supplemen a y and sensi i i y analyses gene ally co obo a ed ou esul s (Supple-
men a y Appendix S1). The associa ion be ween school a endance and TV owne ship—a
placebo exposu e—was mos ly small bu s a is ically signi ican in a ew cases (Supple-
men a y Fig.S10 and TableS18): In Pakis an, gi ls in households wi h a TV had a s a is i-
cally signi ican 7% g ea e school a endance han gi ls ha did no ha e a TV, while boys
wi h a TV had 4% lowe school a endance han boys ha did no . In Tü kiye he asso-
cia ion wi h TV owne ship was subs an ial, nega i e, and s a is ically signi ican , bo h o
boys and gi ls. The associa ion was u he posi i e and s a is ically signi ican o boys in
Colombia and gi ls in A menia.
When s udying yea s o educa ion as an ou come, ins ead o school a endance, Tü kiye
also had by a he la ges associa ion, whe e gi ls wi h a washing machine a home had
a 0.73 (95% CI: 0.54, 0.93) mo e yea s o educa ion han hose wi hou a washe (Sup-
plemen a y Fig.S18). The absolu e di e ence, ob ained om a linea eg ession model,
be ween gi ls wi h and wi hou a washe a home in Tu key was 11 pe cen age poin s (95%
CI: 0.081, 0.15) (Supplemen a y Fig.S12and TableS20).
When adjus ing o household le el ac o s and es ima ing he di e ence in school
a endance be ween boys and gi ls wi hin he same household, ou main esul s we e co -
obo a ed o Tü kiye, whe e gi ls in households wi hou a washe ha e 17% (RR: 0.83;
95% CI: 0.67, 0.70) less school a endance han boys, while in households wi h a washe
gi ls ha e a 9% (RR: 1.09; 95% CI: 0.99, 1.20) ad an age, al hough he la e was no s a-
is ically signi ican (Supplemen a y Fig.S21 and TableS29). In addi ion o Tü kiye, he
in e ac ion e ms o being emale and ha ing a washe was s a is ically signi ican , indi-
ca ing he gi ls we e a a less o a disad an age in households wi h a washe , in Egyp
(95% 1.017, 1.073) and Pakis an (95% CI: 1.004, 1.132), al hough he di e ences we e
a he small. In Albania, gi ls we e a an ad an age in e ms o school a endance, bo h in
households wi h a washe (RR: 1.1;95% CI: 995, 1.224) and wi hou (RR 1.01; 95% CI:
0.98, 1.05): al hough he ad an age was g ea e in households wi h a washe , he in e ac-
ion e ms be ween being emale and ha ing a washe was no s a is ically signician (1.09;
95% CI: 0.98, 1.21).
O he esul s o supplemen a y and sensi i i y analyses a e p esen ed in Appendix S1.
4 Discussion
Using na ionally ep esen a i e da a on 1.6 million adolescen s ages 10–19yea s li ing
in 19 middle-income coun ies, his pape ound no clea ela ionship be ween household
washing machine owne ship and school a endance in mos coun ies. Howe e , in Tü kiye,
young women li ing in a household wi h a washing machine a home had 28% g ea e
school a endance (11 pe cen age poin s) and comple ed 0.73 mo e o al yea s o school-
ing han gi ls who did no ha e access o a washing machine in he household. We also
obse ed a s a is ically signi ican posi i e associa ion be ween washing machine owne -
ship and school a endance o gi ls in Egyp and Albania, al hough much smalle han o
750
Washing machine owne ship and gi ls’ school a endance
1 3
Tü kiye. In Egyp , o example, gi ls wi h a washe a home had 4% g ea e school a end-
ance compa ed o gi ls who did no .
Wi h economic de elopmen and ising incomes, households ha e s a ed acqui ing
household appliances and o he household echnology, which make domes ic wo k mo e
e icien , eeing up ime o o he ac i i ies (B enneman 2014). Since women and gi ls a e
commonly he p ima y ca e ake s o households, hese gains in e iciency can also imp o e
hei oppo uni ies ou side he home in e ms o economic ac i i y and educa ion (G een-
wood e al. 2005). Howe e , al hough appliances can inc ease e iciency, hey do no neces-
sa ily educe he ime spen on household p oduc ion (Ramey and F ancis 2009), and he
ime eed up has been sugges ed o be e-alloca ed o imp o emen s in household hygiene
and di ec ca e o child en (Moky 2000).
Globally, 7.9% o child en 5–17yea s old engage in economic ac i i y (ILO 2022a).
When also conside ing household cho es, his sha e inc eases o 9.5%, acco ding o he
In e na ional Labou O ganiza ion. Howe e , hese indica o s conside “child labo ”
o consis o a leas one hou o economic ac i i y among 5–11-yea -olds, 14h among
12–14-yea -olds, and 43h among 15–14-yea -olds, and household cho es o a leas 21h
o 5–14-yea -olds, while household cho es among 15–17-yea -olds a e no conside ed o
be “child labo ” by he ILO (ILO 2022b). The e o e, equen wo k a an ea ly age, includ-
ing household cho es, can po en ially comp omise he educa ional de elopmen o a highe
sha e o child en han indica ed by he ILO child labo s a is ics.
Fo boys, he associa ion be ween washing machine owne ship and school a endance
was small and mos ly no s a is ically signi ican in all coun ies. Indeed, boys gene ally
engage mo e in child labo , pa icula ly in amily wo k (e.g., on a amily a m) and wo k
ou side he home, while gi ls usually ha e been ound o do mo e household wo k (i.e.,
cho es), mi o ing commonly obse ed gende no ms (Pu nick and Bo ns ein 2016). Wi h-
ou mode n home appliances and u ili ies, household wo k can ake o e 50h pe week,
wi h an addi ional 15–30h when he e is an in an o ca e o in he household (Ramey and
F ancis 2009). In low- and middle-income coun ies, mos households do no own appli-
ances like washing machines and many do no ha e access o piped wa e o elec ici y.
Laund y may need o be washed by hand a e e ching wa e o a eling o a wa e sou ce
in he mo nings (du ing school hou s) o be d y and eady by nigh ime (Assaad e al.
2010).
The ew cases o posi i e ela ionship be ween washing machine owne ship and gi ls’
school a endance also appea ed o be pe iod dependen and he associa ion a ied ac oss
su ey yea s wi hin coun ies. Fo example, he posi i e associa ion was no obse ed in
he mos ecen su eys in Tü kiye (2013), Albania (2017–18), and Egyp (2014), when
school a endance and washing machine owne ship a e we e e y high, pa icula ly in Tü -
kiye and Albania, and he e o e li le a iance o iden i y a ela ionship. Mo eo e , school
policies in ecen decades ha e inc easingly mo ed o aising he compulso y schooling age
and u he en o cing hese policies (e.g., h ough penal ies and, in ex emis, sho - e m
imp isonmen o pa en s who do no send hei child en o school), which may inc ease he
oppo uni y cos s o keeping adolescen s a home. Fo example, Tü kiye inc eased compul-
so y schooling om 8 o 12yea s in 2012.
E en hough ou esul s indica ed ha a ela ionship be ween washing machine owne -
ship and school a endance was no a gene al phenomenon, he ela ionship we obse ed in
Tü kiye may be eal and based on speci ic a ibu es o he Tu kish con ex , o example,
gende no ms. The e may be speci ic s uc u al, economic, o cul u al ac o s p esen which
make washing machine owne ship bene icial o gi l’s school a endance. In his pape , we
we e unable o explo e hese ac o s.
751
O.Ka lsson, J.-W.De Ne e
1 3
4.1 S udy limi a ions
Despi e ou e o o de e mine he ela ionship be ween household owne ship o a washing
machine and adolescen s’ school a endance, ou s udy has se e al limi a ions.
Fi s , washing machine owne ship is no andom, and he es ima ed ela ionship be ween
washe owne ship and school a endance may su e om con ounding. The inhe en link
be ween washe owne ship and li ing s anda ds and socioeconomic s a us—which in u n
imp o es school a endance—is clea . Howe e , we con olled o se e al measu es o li ing
s anda ds and socioeconomic s a us, al hough he impac o some esidual con ounding canno
be uled ou . Fu he , he pa e n obse ed o washe owne ship was gene ally no obse ed
o owne ship o o he asse s, which we e less likely o a ec schooling bu also inhe en ly
linked o li ing s anda ds, p ima ily TV owne ship.
Pa en s wi h a s ong p e e ence o hei child o a end school may also be mo e likely o
p io i ize ge ing a washing machine o educe he bu den o child labo , as well as engaging
in o he beha io s ha inc ease he child’s school a endance and he eby bias he ela ionship
upwa ds. Despi e a po en ial upwa d con ounding bias om li ing s anda ds and pa en al p e -
e ences, we do no ind a gene al e ec o washe owne ship on school a endance.
Second, he owne ship o a washing machine signals li ing s anda ds a yingly, depend-
ing on he con ex . In places whe e owne ship is close o uni e sal, non-owne ship is mo e
likely o signal ela i e disad an age. Whe e educa ion is nea ly uni e sal, non-a endance is
expec ed o e lec ela i e disad an ages o a g ea e ex en han whe e o e all school a end-
ance is less common. We es ic ed ou main analyses o coun ies wi h be ween 5 and 95%
washe owne ship and school a endance (wi h he a e age ac oss coun ies being 48% washe
owne ship and 82% school a endance) o ensu e su icien a iance o ob aining es ima es.
Howe e , he ex en o sa u a ion (i.e., comple e p esence o absence) in washe owne ship
and school a endance should be kep in mind, especially in sub-analyses (e.g., when s udying
he ela ionships ac oss ime o household weal h).
Thi d, we did no ha e da a on child en’s ime use, which would ha e allowed us o exam-
ine he associa ion be ween domes ic and economic ac i i ies and schooling ou comes mo e
di ec ly. Fo example, al hough we gene ally do no ind an associa ion wi h school a end-
ance, washe owne ship may ha e ins ead inc eased economic ac i i y ou side he home
among gi ls. Fu he , acqui ing a washe may in some cases ha e enabled adolescen s o do
mo e household wo k, o example i he amily swi ched om using laund y se ices o
doing laund y hemsel es (Cowan 1983). Fou h, ou p ima y ou come indica ed any school
a endance in he yea o he su ey bu con ained li le in o ma ion abou he consis ency o
a endance h oughou he school yea . Howe e , ou esul s we e simila when using he o al
numbe o yea s o educa ion comple ed as ou ou come. Fi h, al hough unlikely, ou models
a e ulne able o e e se causali y because o po en ial misalignmen in he iming be ween
ou exposu e and ou come. Fo example, i child en lea ned abou he bene i s o washing
machines a school and induced hei households o pu chase hem.
5 Conclusions
This pape ound no clea link be ween washing machine owne ship and school a endance
o ei he gi ls o boys ac oss se e al middle-income coun ies. Howe e , in Tü kiye, gi ls
wi h a washing machine a home we e subs an ially mo e likely o a end school and had
comple ed mo e o al yea s o schooling han gi ls wi hou a washing machine a home. No
752
Washing machine owne ship and gi ls’ school a endance
1 3
such di e ences we e obse ed o boys. Simila ly, gi ls in Egyp and Albania appea ed
o be mo e likely o a end school when hey had a washe a home, al hough he links
we e much smalle han in Tü kiye. While washing machine owne ship does no gene ally
imp o e gi ls’ school a endance, i may do so in some con ex s.
The Uni ed Na ions’ Sus ainable De elopmen Goals mo e beyond p ima y educa ion
o ocus on achie ing uni e sal seconda y educa ion. To achie e uni e sal seconda y edu-
ca ion, policymake s should conside labo -sa ing household echnologies, which educe
he oppo uni y cos o amilies o send hei child en o school. Fu he s udies could es
he causal e ec s o household owne ship o washing machines and imp o ing access o,
o example, public laund oma s o u he imp o e gi ls’ school pe o mance.
Supplemen a y In o ma ion The online e sion con ains supplemen a y ma e ial a ailable a h ps:// doi.
o g/ 10. 1007/ s10888- 023- 09612-7.
Acknowledgemen s Ka lsson was unded by Ve enskaps ade (2019-06396).
Au ho con ibu ions Ka lsson did da a managemen , analysis, and epo ing and led he w i ing o he
manusc ip . De Ne e con ibu ed o he w i ing o he manusc ip and p o ided c i ical e iew o he s udy.
Funding Open access unding p o ided by Lund Uni e si y.
Da a a ailabili y All Demog aphic and Heal h Su eys a e a ailable a h ps:// dhsp og am. com ( equi ing a
simple applica ion).
Decla a ions
This p ojec used publicly accessible seconda y da a om he Demog aphic and Heal h Su eys websi e.
These ac i i ies did no mee he egula o y de ini ion o human subjec esea ch.
Decla a ions o in e es None.
Compe ing in e es s The au ho s decla e no compe ing in e es s.
Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long
as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Com-
mons licence, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his a icle
a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line o he
ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended use is no
pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly
om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
Re e ences
Aliaga, A., Ren, R.: Op imal sample sizes o wo-s age clus e sampling in Demog aphic and Heal h Su eys.
DHS Wo king Pape s. h ps:// www. dhsp og am. com/ pubs/ pd / WP30/ WP30. pd (2006)
Assaad, R., Le ison, D., Zibani, N.: The e ec o domes ic wo k on gi ls’ schooling: e idence om Egyp . Fem.
Econ. 16(1), 79–128 (2010). h ps:// doi. o g/ 10. 1080/ 13545 70090 33827 29
Bha ga a, S., & Ke , A.: The e ec o ime-sa ing household appliance owne ship on ou comes o child en
and ma ied women: e idence om India. IZA J. Labo Policy, 12(1), (2021). h ps:// doi. o g/ 10. 2478/ izajo
lp- 2022- 0008
B enneman, A.: In as uc u e & po e y linkages: a li e a u e e iew, 2002. Wo ld Bank, Washing on (2014)
Caldwell, J.C.: Educa ion as a ac o in mo ali y decline an examina ion o nige ian da a. Popul S ud 33(3),
395–413 (1979). h ps:// doi. o g/ 10. 2307/ 21738 88
Co si, D.J., Neuman, M., Finlay, J.E., Sub amanian, S.V.: Demog aphic and heal h su eys: a p o ile. In . J.
Epidemiol. 41(6), 1602–1613 (2012)
753
O.Ka lsson, J.-W.De Ne e
1 3
Cowan, R.S.: Mo e wo k o mo he : The i onies o household echnology om he open hea h o he mic o-
wa e. Basic Books, New Yo k (1983)
DHS: Publica ions by Coun y. The DHS P og am. h p:// www. dhsp og am. com/ Publi ca io ns/ Publi ca io ns- by-
Coun y. c m (2022). Accessed 30 May 2022
G ae z, N., Woyczynski, L., Wilson, K. F., Hall, J. B., Aba e, K. H., Abd-Allah, F., Adebayo, O. M., Ade-
kanmbi, V., A sha i, M., Ajumobi, O., Akinyemiju, T., Alahdab, F., Al-Aly, Z., Rabanal, J. E. A., Ali-
janzadeh, M., Alipou , V., Al i kawi, K., Ami esmaili, M., Anbe , N. H., …, Local Bu den o Disease
Educa ional A ainmen Collabo a o s: Mapping dispa i ies in educa ion ac oss low- and middle-income
coun ies. Na u e. 577(7789), A icle 7789 (2020). h ps:// doi. o g/ 10. 1038/ s41586- 019- 1872-1
G eenwood, J., Seshad i, A., Yo ukoglu, M.: Engines o libe a ion. Re . Econ. S ud. 72(1), 109–133 (2005)
Howe, L.D., Galoba des, B., Ma ijase ich, A., Go don, D., Johns on, D., Onwujekwe, O., Pa el, R., Webb, E.A.,
Lawlo , D.A., Ha g ea es, J.R.: Measu ing socio-economic posi ion o epidemiological s udies in low-
and middle-income coun ies: a me hods o measu emen in epidemiology pape . In . J. Epidemiol. 41(3),
871–886 (2012). h ps:// doi. o g/ 10. 1093/ ije/ dys037
ILO: S a is ics on child labou . ILOSTAT. h ps:// ilos a . ilo. o g/ opics/ child- labou /# (2022a). Accessed 23 Oc
2022
ILO: Wha is child labou . In e na ional Labou O ganiza ion. h ps:// www. ilo. o g/ ipec/ ac s/ lang-- en/ index. h m
(2022b). Accessed 9 Dec 2022
IPUMS: Desc ip ion: WASHERYN (WASHERYN). IPUMS DHS. h ps:// www. idhsd a a. o g/ idhs- ac ion/ a ia
bles/ WASHE RYN# desc ip ion_ sec i on (2022). Accessed 6 Dec 2022
Jabba ian, J., We ne , L.K., Kagoné, M., Lemp, J.M., McMahon, S., Ho s ick, O., Kazianga, H., Kobiané, J.-F.,
Fink, G., Ne e, J.-W.D.: “I ’s he po e y”—S akeholde pe spec i es on ba ie s o seconda y educa ion
in u al Bu kina Faso. PLoS ONE 17(11), e0277822 (2022). h ps:// doi. o g/ 10. 1371/ jou n al. pone. 02778 22
Ka lsson, O., De Ne e, J.-W., Sub amanian, S.V.: Weakening associa ion o pa en al educa ion: analysis o child
heal h ou comes in 43 low- and middle-income coun ies. In . J. Epidemiol. 48(1), 83–97 (2018). h ps://
doi. o g/ 10. 1093/ ije/ dyy158
Ka lsson, O., Kim, R., Joe, W., Sub amanian, S.: The ela ionship o household asse s and ameni ies wi h child
heal h ou comes: an explo a o y c oss-sec ional s udy in India 2015–2016. SSM-Popul. Heal h 10, 100513
(2020)
Ke , A.: Household in es men in du able appliances and ou comes o child en: e idence om China. Labou
Econ. 58, 110–127 (2019). h ps:// doi. o g/ 10. 1016/j. labeco. 2019. 04. 010
Moky , J.: Why “mo e wo k o mo he ?” Knowledge and household beha io , 1870–1945. J. Econ. His . 60(1),
1–41 (2000)
De Ne e, J.-W., Fink, G., Sub amanian, S.V., Moyo, S., Bo , J.: Leng h o seconda y schooling and isk o HIV
in ec ion in Bo swana: e idence om a na u al expe imen . Lance Glob. Heal h 3(8), e470–e477 (2015).
h ps:// doi. o g/ 10. 1016/ S2214- 109X(15) 00087-X
De Ne e, J.-W., Ha gen, K., Ve gue , S.: Na ionally and egionally ep esen a i e analysis o 1.65 million chil-
d en aged unde 5 yea s using a child-based human de elopmen index: a mul i-coun y c oss-sec ional
s udy. PLOS Med. 17(3), e1003054 (2020). h ps:// doi. o g/ 10. 1371/ jou n al. pmed. 10030 54
Pu nick, D.L., Bo ns ein, M.H.: Is child labo a ba ie o school en ollmen in low- and middle-income coun-
ies? In . J. Educ. De . 41, 112–120 (2015). h ps:// doi. o g/ 10. 1016/j. ijedu de . 2015. 02. 001
Pu nick, D.L., Bo ns ein, M.H.: Vi. gi ls’ and boys’ labo and household cho es in low- and middle-income
coun ies. Monog . Soc. Res. Child De . 81(1), 104–122 (2016). h ps:// doi. o g/ 10. 1111/ mono. 12228
Ramey, V.A., F ancis, N.: A cen u y o wo k and leisu e. Am. Econ. J.-Mac oecon. 1(2), 189–224 (2009).
h ps:// doi. o g/ 10. 1257/ mac.1. 2. 189
Schunck, R., Pe ales, F.: Wi hin-and be ween-clus e e ec s in gene alized linea mixed models: a discussion o
app oaches and he x hyb id command. S and. Genomic Sci. 17(1), 89–115 (2017)
Shi, J., Wang, F., Wang, H.: The e ec o household echnology on child heal h: e idence om China’s “home
appliances going o he coun yside” policy. In . J. En i on. Res. Public Heal h 19(19), A icle 19 (2022).
h ps:// doi. o g/ 10. 3390/ ije p h1919 11976
UNESCO: Da a o he Sus ainable De elopmen Goals. UNESCO Ins i u e o S a is ics.h ps:// uis. unesco.
o g/(2022). Accessed 21 Oc 2022
WHO: Heal h in 2015: om MDGs o SDGs. Gene a: Wo ld Heal h O ganiza ion. h ps:// i is. who. in / bi s
eam/ handle/ 10665/ 200009/ 97892 41565 110_ eng. pd ? seque nce=1 (2015). Accessed 17 Dec 2023
WHO: Adolescen heal h. Wo ld Heal h O ganiza ion. h ps:// www. who. in / heal h- opics/ adole scen - heal h#
ab= ab_1 (2022). Accessed 6 Dec 2022
Woold idge, J.M.: Co ela ed andom e ec s models wi h unbalanced panels. J. Econom. 211(1), 137–150
(2019). h ps:// doi. o g/ 10. 1016/j. jecon om. 2018. 12. 010
Zou, G.: A modi ied poisson eg ession app oach o p ospec i e s udies wi h bina y da a. Am. J. Epidemiol.
159(7), 702–706 (2004)
754

Washing machine owne ship and gi ls’ school a endance
1 3
Zou, G., Donne , A.: Ex ension o he modi ied poisson eg ession model o p ospec i e s udies wi h co ela ed
bina y da a. S a . Me hods Med. Res. 22(6), 661–670 (2013). h ps:// doi. o g/ 10. 1177/ 09622 80211 427759
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and
ins i u ional a ilia ions.
755