Balza, Lenin H.; Cua as, Jo ge; Gomez-Pa a, Nicolas; Se eb isky, Tomás
Wo king Pape
In as uc u e se ices and ea ly childhood de elopmen
in La in Ame ica and he Ca ibbean: Wa e , sani a ion, and
ga bage collec ion
IDB Wo king Pape Se ies, No. IDB-WP-1612
P o ided in Coope a ion wi h:
In e -Ame ican De elopmen Bank (IDB), Washing on, DC
Sugges ed Ci a ion: Balza, Lenin H.; Cua as, Jo ge; Gomez-Pa a, Nicolas; Se eb isky, Tomás (2024) :
In as uc u e se ices and ea ly childhood de elopmen in La in Ame ica and he Ca ibbean:
Wa e , sani a ion, and ga bage collec ion, IDB Wo king Pape Se ies, No. IDB-WP-1612, In e -
Ame ican De elopmen Bank (IDB), Washing on, DC,
h ps://doi.o g/10.18235/0012998
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/299453
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/3.0/igo/
In as uc u e Se ices and Ea ly Childhood
De elopmen in La in Ame ica and he
Ca ibbean: Wa e , Sani a ion, and Ga bage
Collec ion
Lenin H. Balza
Jo ge Cua as
Nicolas Gomez-Pa a
Tomás Se eb isky
WORKING PAPER No IDB-WP-1612
In e -Ame ican De elopmen Bank
In as uc u e and Ene gy Sec o
June 2024
In as uc u e Se ices and Ea ly Childhood
De elopmen in La in Ame ica and he
Ca ibbean: Wa e , Sani a ion, and Ga bage
Collec ion
Lenin H. Balza
Jo ge Cua as
Nicolas Gomez-Pa a
Tomás Se eb isky
In e -Ame ican De elopmen Bank
In as uc u e and Ene gy Sec o
June 2024
Ca aloging-in-Publica ion da a p o ided by he
In e -Ame ican De elopmen Bank
Felipe He e a Lib a y
In as uc u e se ices and ea ly childhood de elopmen in La in Ame ica and he Ca ibbean: wa e ,
sani a ion, and ga bage collec ion / Lenin H. Balza, Jo ge Cua as, Nicolas Gomez-Pa a, Tomás
Se eb isky.
p. cm. — (IDB Wo king Pape Se ies; 1612)
Includes bibliog aphical e e ences.
1. In as uc u e (Economics)-La in Ame ica. 2. In as uc u e (Economics)-Ca ibbean A ea. 3. Wa e
supply-La in Ame ica. 4. Wa e supply-Ca ibbean A ea. 5. Sani a ion-La in Ame ica. 6. Sani a ion-
Ca ibbean A ea. 7. Re use and e use disposal-La in Ame ica. 8.
Re use and e use disposal-Ca ibbean A ea. 9. Toddle s-Heal h and hygiene-La in Ame ica. 10.
Toddle s-Heal h and hygiene-Ca ibbean A ea. I. Balza, Lenin. II. Cua as, Jo ge. III. Gomez-Pa a,
Nicolas. IV. Se eb isky, Tomás. V. In e -Ame ican De elopmen Bank. In as uc u e and Ene gy
Sec o . VI. Se ies.
IDB-WP-1612
h p://www.iadb.o g
Copy igh © 2024 In e -Ame ican De elopmen Bank ("IDB"). This wo k is subjec o a C ea i e
Commons license CC BY 3.0 IGO (h ps://c ea i ecommons.o g/licenses/by/3.0/igo/legalcode). The
e ms and condi ions indica ed in he URL link mus be me and he espec i e ecogni ion mus be
g an ed o he IDB.
Fu he o sec ion 8 o he abo e license, any media ion ela ing o dispu es a ising unde such license
shall be conduc ed in acco dance wi h he WIPO Media ion Rules. Any dispu e ela ed o he use o
he wo ks o he IDB ha canno be se led amicably shall be submi ed o a bi a ion pu suan o he
Uni ed Na ions Commission on In e na ional T ade Law (UNCITRAL) ules. The use o he IDB's name
o any pu pose o he han o a ibu ion, and he use o IDB's logo shall be subjec o a sepa a e
w i en license ag eemen be ween he IDB and he use and is no au ho ized as pa o his license.
No e ha he URL link includes e ms and condi ions ha a e an in eg al pa o his license.
The opinions exp essed in his wo k a e hose o he au ho s and do no necessa ily e lec he iews o
he In e -Ame ican De elopmen Bank, i s Boa d o Di ec o s, o he coun ies hey ep esen .
In as uc u e Se ices and Ea ly Childhood
De elopmen in La in Ame ica and he
Ca ibbean: Wa e , Sani a ion, and Ga bage
Collec ion∗
Lenin H. Balza Jo ge Cua as Nicolas Gomez-Pa a Tom´as Se eb isky
June 2024
Abs ac
Access o essen ial in as uc u e se ices such as wa e , sani a ion, and ga bage collec ion
can conside ably a ec child en’s en i onmen and may play a signi ican ole in shaping ea ly
childhood de elopmen al and heal h ou comes. Using da a om he Mul iple Indica o Clus-
e Su eys (MICS) and he Demog aphic and Heal h Su eys (DHS) o 18 coun ies in La in
Ame ica and he Ca ibbean (LAC), we show a signi ican posi i e associa ion be ween access o
wa e and sani a ion and ea ly childhood de elopmen , as well as educed ins ances o s un ing.
In addi ion, we iden i y a nega i e associa ion be ween access o imp o ed ga bage collec ion
se ices and he a es o s un ing and unde weigh among child en unde i e. Ou indings
a e obus a e using al e na i e measu es o access and con olling o indi idual, ma e nal,
and household ac o s, alongside conside a ions o household weal h and ca egi e ’s s imula ion
ac i i ies. Simila ly, he economic ele ance o he ela ionship is highligh ed by he subs an ial
gap ela i e o he size o he ulne able g oups, pe sis ing e en a e adjus ing o con ounding
a iables. Ou esul s also sugges ha households may be able o lessen he po en ial impac o
pollu an s h ough mi iga ion measu es such as ea ing wa e o make i sa e o consump ion,
using handwashing cleanse s, and s o ing household ash in lidded con aine s. The cu en
indings unde sco e he impo ance o in es ing in basic in as uc u e se ices as a c i ical com-
ponen o comp ehensi e s a egies o enhance ea ly childhood de elopmen and heal h in low-
and middle-income coun ies. We emphasize he impo ance o conside ing he quali y and ype
o in as uc u e se ices alongside hei a ailabili y. Fu u e esea ch should inco po a e mo e
comple e and de ailed da a o imp o e unde s anding o he causal ela ionship be ween wa e ,
sani a ion, and ga bage collec ion and ea ly childhood de elopmen , as well as he mechanisms
unde lying he obse ed associa ions.
JEL codes: H41, I15, J13, Q53.
Keywo ds: Ea ly childhood de elopmen , Wa e and sani a ion access, Was e managemen
in as uc u e, Low- and middle-income coun ies, Mi iga ion s a egies.
∗Balza: In e -Ame ican De elopmen Bank (lenin[email p o ec ed]); Cua as: Ha a d G adua e School o Educa-
ion/Uni e sidad de los Andes ([email p o ec ed]a d.edu); Gomez-Pa a: In e -Ame ican De elopmen Bank (ngomez-
[email p o ec ed]); Se eb isky: In e -Ame ican De elopmen Bank ([email p o ec ed]). The indings, in e p e a-
ions, and conclusions exp essed in his wo k do no necessa ily e lec he iews o he In e -Ame ican De elopmen
Bank, hei Boa ds o Execu i e Di ec o s, o he go e nmen s hey ep esen . We hank Lisa Bagnoli, Daniela
Collazos, Giuliana Daga, Flo encia Lopez Boo, Da id Ma ´ıas, Ca ol Te acina Ha man, and an anonymous e iewe
o hei commen s and sugges ions. Any e o s o omissions emain he sole esponsibili y o he au ho s.
1 In oduc ion
Mo e han 250 million child en younge han i e li ing in low- and middle-income coun ies a e
a isk o no eaching hei de elopmen al po en ial due o ad e si y and lack o nu u ing ca e
(i.e., good heal h, adequa e nu i ion, ea ly lea ning oppo uni ies, sa e y and secu i y, and espon-
si e ca egi ing) (Black e al.,2017;Lu e al.,2020;McCoy e al.,2022). This is conce ning as
ea ly childhood de elopmen (ECD) s ongly p edic s educa ional ajec o ies, u u e heal h, and
economic and social well-being h oughou an indi idual’s li espan (Black e al.,2017;Heckman,
2006). Indeed, b ain de elopmen and he acquisi ion o ounda ional skills a e highly sensi i e
o en i onmen al inpu ea ly in li e (She idan and Nelson,2009). Ea ly expe iences – posi i e o
nega i e – hus ha e eno mous long- e m consequences o indi idual and socie al wel a e (Black
e al.,2017;Cunha and Heckman,2007;Knudsen e al.,2006).
Ex ensi e e idence has shown ha ECD is pa icula ly sensi i e o a ious nu i ional, heal h,
and psychosocial ac o s, including he home en i onmen , access o lea ning oppo uni ies, and
exposu e o iolence (A anasio e al.,2020;Cua as,2021,2022;Cua as e al.,2020,2021;Jeong
e al.,2017;McCoy e al.,2016;Salhi e al.,2021;Wol and McCoy,2019). While he e is ex ensi e
li e a u e wi hin he ield o public heal h assessing he impac o he physical en i onmen (in
pa icula , he ela ion be ween wa e , sani a ion, and hygiene wi h nu i ion) on ECD (Danaei
e al.,2016;Sania e al.,2019), he bulk o his esea ch has ocused on s un ing as an ou come
and no child de elopmen speci ically. S un ing has o en been used as a p oxy o child de elop-
men , hough ecen e idence sugges s ha i is an inadequa e p oxy (Sania e al.,2019;Pe umal
e al.,2018). In his con ex , ela i ely li le e o has been made o in eg a e de elopmen al science
wi h en i onmen al heal h (T en acos a e al.,2016). This wo k is none heless impo an as clima e
change, indus ializa ion and poo basic in as uc u e access ha e inc eased se e al speci ic en i-
onmen al isks ha may be pa icula ly in luen ial ea ly in li e. These a e especially p onounced
in low- and middle-income coun ies (LMICs), whe e millions o child en lack access o clean wa-
e sou ces, imp o ed sani a ion, and adequa e was e managemen (McCoy e al.,2022). Indeed,
poo wa e , sani a ion, and ga bage disposal ep esen pa icula isks o child en as hey consume
mo e wa e , ood, and ai ela i e o hei body weigh han adul s, while hei me abolism is less
capable o coping wi h con aminan s and hei b ain a chi ec u e is mo e sensi i e o pollu an s
(T en acos a e al.,2016).
Inadequa e o pollu ed wa e , poo sani a ion, and ga bage disposal can a ec ECD h ough
a leas h ee mechanisms. Fi s , poo hygiene, pollu ed wa e , and inadequa e sani a ion and
was e managemen can lead o biological changes a ec ing he me abolic sys em wi h downs eam
consequences on child en’s weigh and nu i ion (Heindel and Schug,2014). Second, inges ing
pollu an s and mic obial con aminan s can dis up he no mal de elopmen and p uning o synapses
in key a eas o he b ain (e.g., he p e on al co ex), impac ing b ain s uc u e and unc ion
and consequen ly cogni ion, social-emo ional de elopmen , and beha io (Walke e al.,2011).
1
Handwashing and combined wa e , sani a ion, handwashing, and nu i ion in e en ions migh
imp o e child mo o de elopmen (S ewa e al.,2018).1Finally, pollu ion and he lack o clean
wa e , adequa e sani a ion, and was e managemen sou ces can be a majo sou ce o s ess o
amilies, which can a ec hei beha io and he ype and quali y o ca e ha hey p o ide o
child en (Bo ns ein e al.,2015). In sum, e idence suppo s hypo hesizing ha inadequa e wa e ,
sani a ion, and ga bage disposal can impac childhood heal h and de elopmen , wi h downs eam
consequences o li elong indi idual and socie al ou comes.
Consis en wi h hese indings, in es ing in imp o ed wa e , sani a ion, and was e managemen
in as uc u e can be a key s a egy o p omo ing posi i e heal h and de elopmen al ou comes
ea ly in li e (Acu and Ka ine,2013). En i onmen al p oblems such as odo s and ai and wa e
pollu ion a e caused by poo was e managemen and may be exace ba ed as he clima e c isis
deepens (Janmaimool,2017;Vacca i e al.,2019). Poo was e managemen inc eases he isk
o se e e diseases, like dysen e y and jaundice (Ra hnamala e al.,2021). Howe e , budge and
inancial cons ain s, socioeconomic ac o s, and nega i e a i udes owa ds was e managemen
make i di icul o in es in was e managemen in as uc u e in LMICs (Al-Kha ib e al.,2016;
Vacca i e al.,2019;S¸chiopu e al.,2007;Manga e al.,2019;Wang e al.,2018;Kese e al.,2012;
Ra hnamala e al.,2021). Fu he mo e, hese policies a e ha d o a ge as esponsibili y o he
gene a ion o solid was e is sha ed ac oss indus y, households, and go e nmen (Abdul edha e al.,
2018;Yamamo o and Kinnaman,2022;Hamil on e al.,2013;Agyei-Mensah and O eng-Ababio,
2012).
Add essing en i onmen al in as uc u e gaps is c ucial o enhancing ECD in La in Ame ica and
he Ca ibbean (LAC) egion (G¨un he and Laho i,2021;Schady,2015;Ko c and Hauchman,2021).
While LAC has made signi ican s ides in enhancing access o clean wa e , sani a ion emains a
p essing conce n, especially in u al a eas (Al a ez,2019;Jeuland e al.,2013). Simila ly, he egion
has made signi ican p og ess in was e managemen , wi h an a e age solid was e collec ion a e
highe han he global a e age. Ye , challenges pe sis in achie ing consis en and equi able was e
managemen se ices ac oss di e en a eas. The egion g apples wi h a ia ions in se ice quali y
and equency, as well as a low le el o municipal planning in solid was e managemen (G au e al.,
2015). These in as uc u al challenges no only pose immedia e heal h isks, bu also could ha e
long- e m implica ions on cogni i e and physical de elopmen . The egion’s unique socio-economic
dynamics, coupled wi h i s di e se opog aphy and clima e, u he accen ua e hese challenges.
Vulne able a eas o en a e cha ac e ized by limi ed s a e p esence and in es men . Families in
hese loca ions o en lack he esou ces o knowledge o adop mi iga i e measu es, exace ba ing
he heal h and de elopmen al isks o child en. In s a k con as , u ban cen e s, despi e be e
in as uc u e, g apple wi h challenges posed by apid u baniza ion and popula ion densi y, o en
leading o o e bu dened acili ies (Spa kman and S u zenegge ,2016).
1Howe e , modes ly imp o ed sani a ion p ac ices do no necessa ily enhance child heal h ou comes, unde sco ing
he complexi y o achie ing heal h bene i s h ough in e en ions (Pa il e al.,2014).
2
This pape aims o p o ide e idence on he po en ial posi i e e ec s o imp o ed wa e , sani a ion,
and was e managemen in as uc u e on ECD in LAC. Ou indings can help p o ide impo an
e idence o suppo a ge ed in es men s ha bene i he en i onmen , local popula ions, and
young child en. In pa icula , we seek o add ess he ollowing esea ch ques ions. Fi s , does
access o imp o ed wa e , sani a ion, and was e managemen impac young child en’s heal h and
de elopmen ? Second, o wha ex en do he links be ween access o imp o ed wa e , sani a ion, and
was e disposal a y ac oss coun ies, when con olling o indi idual and household cha ac e is ics,
such as ma e nal educa ion, weal h, heal h s a us, s imula ion beha io s, and he a ailabili y o
ce ain asse s and goods (e.g., hand cleanse s, e ige a o s, access o bo led wa e , and lidded
con aine s)?
We use wo main da a sou ces o examine hese ques ions. Fi s , we employ UNICEF’s Mul iple
Indica o s Clus e Su eys (MICS) om Belize, Cos a Rica, Cuba, El Sal ado , Guyana, Hondu as,
Jamaica, Panama, Pa aguay, Dominican Republic, Sain Lucia, Su iname, T inidad and Tobago,
and U uguay, which include in o ma ion on access o imp o ed wa e and sani a ion and on child
heal h, and de elopmen . We also use he Demog aphic and Heal h Su eys (DHS) om Boli ia,
Colombia, Dominican Republic, Guyana, Nica agua, and Pe u, which p o ide in o ma ion on child
heal h and nu i ion, and household was e managemen . We conduc ed a ho ough mul i a ia e
s a is ical analysis, u ilizing linea p obabili y models o explo e he co ela ions be ween in as-
uc u e access and de elopmen al and heal h ou comes, while aking ad an age o he di e ences
in in as uc u e be ween coun ies. To mi iga e po en ial endogenei y conce ns, especially he con-
ounding e ec o household weal h, we inco po a ed con ol a iables in ou models. We u he
alida ed ou indings by inco po a ing he s imula ion index, which is a widely ecognized p edic-
o o ECD. Addi ionally, we used ca ego ical a iables o analyze access o wa e and sani a ion,
and applied a p opensi y sco e ma ching app oach o he da a. Despi e hese a ia ions, ou esul s
emain consis en .
The indings o ou pape signi ican ly emphasize he co e associa ions be ween access o im-
p o ed wa e , sani a ion, was e managemen , and he de elopmen al and heal h ou comes in young
child en. Child en in households wi h be e access o hese acili ies a e mo e likely o be de el-
opmen ally on ack and less p one o s un ing. No able a iabili y in hese associa ions ac oss
coun ies calls o u he explo a ion o o he en i onmen al ac o s a ec ing child en’s heal h and
de elopmen . Addi ionally, imp o ed ga bage collec ion se ices a e linked wi h educed a es o
s un ing and unde weigh among child en unde i e. We also explo e mi iga ing beha io s ha
households can adop amid inadequa e access o clean wa e and sani a ion acili ies. Ou indings
ad oca e o p omo ing household wa e ea men p ac ices and using lidded con aine s o was e
as p o isional solu ions owa ds uni e sal access o wa e and sani a ion. Examining p edic ed
a es o s un ing and unde weigh among child en who ha e expe ienced ecen illness episodes
highligh s po en ial unde lying mechanisms a ec ing child de elopmen , emphasizing he heal h
channel.
3
Ou indings esona e wi h and ex end he exis ing li e a u e, ma king an ad ancemen in com-
p ehending he in as uc u e-heal h nexus. Gao e al. (2021), using a sample o mo e han 88,000
child en aged 26-59 mon hs in 20 coun ies in Sub-Saha an A ica, ound a posi i e associa ion
be ween a composi e index o imp o ed housing (including access o be e wa e sou ces and
sani a ion acili ies) and ea ly childhood de elopmen . In addi ion, Schady (2015) highligh s he
dispa i ies in wa e and sani a ion in as uc u e access in LAC. I ou lines he public heal h im-
plica ions o inadequa e acili ies, linking mic obial inges ion o heal h ad e si ies such as dia hea,
a signi ican ac o in unde - i e mo ali y in de eloping egions. While exis ing e idence subs an-
ia es he ela ionship be ween imp o ed in as uc u e and educed child mo ali y, a u he need
exis s o conduc new e alua ions o assess he impac on child mo bidi y, nu i ional s a us, and
de elopmen . The e is a lack o well-iden i ied e idence wi hin he LAC egion, ad oca ing o
comp ehensi e e alua ions o unde s and he mul i-dimensional bene i s o imp o ed wa e and
sani a ion access o child en o e ime (Schady,2015;G¨un he and Laho i,2021). To he bes
o ou knowledge, ou pape is he i s o explo e he independen ela ionships be ween access
o wa e , sani a ion, and was e managemen , and hei impac on ECD and heal h ou comes in
LAC using be ween-coun y he e ogenei y. We pa icula ly emphasize he inclusion o was e man-
agemen as pa o access o basic in as uc u e. This disagg ega ed in o ma ion is c i ical o
unde s anding speci ic ea ly childhood de elopmen al and heal h ou comes ha may be impac ed
by physical en i onmen ea u es and o in o ming u u e in es men s in in as uc u e o p omo e
posi i e de elopmen al ajec o ies om a young age.
The emainde o his pape is o ganized as ollows. In sec ion 2we summa ize he me hodology,
including he measu emen ools and es ima ion me hod. Sec ion 3desc ibes he da a used o
assess he links be ween access o imp o ed wa e sou ces, sani a ion, and ga bage disposal and
ea ly childhood de elopmen al and heal h ou comes. Sec ion 4summa izes he esul s. We conclude
wi h a gene al discussion and implica ions o u u e esea ch and p ac ice.
2 Me hodology: Measu emen and es ima ion
In his pape , we conside wo main ou come a iables: (1) child de elopmen and (2) child
heal h/nu i ion. We use he ea ly childhood de elopmen index (ECDI) (Loizillon e al.,2017) as
a measu e o young child en’s de elopmen . The ECDI is a 10-i em pa en - epo ed scale ha aims
o assess he basic de elopmen al skills o child en aged 3-5 yea s in ou de elopmen al domains:
social-emo ional, li e acy-nume acy, app oaches o lea ning, and physical. The social-emo ional
domain consis s o h ee i ems asking whe he he child (1) ge s along well wi h o he child en, (2)
kicks, bi es, o hi s o he child en o adul s, and (3) is easily dis ac ed. Two i ems co e li e acy-
nume acy: (1) whe he he child can iden i y o name a leas en le e s o he alphabe , and (2)
whe he he child can ead a leas ou simple wo ds. App oaches o lea ning is measu ed by wo
i ems: (1) whe he he child ollows simple ins uc ions and (2) whe he , when gi en some hing o
4
Figu e 4: Access o wa e and sani a ion in as uc u e by su ey
(a) Wa e access (dummy)
BZ2006 (77.4%) BZ2011 (84.0%) BZ2015 (87.5%) CR2011 (99.2%) CU2019 (98.2%)
DO2014 (95.1%) DO2019 (96.9%) GY2006 (50.9%) GY2014 (69.0%) GY2019 (70.0%)
HN2019 (92.7%) JM2011 (89.6%) LC2012 (97.3%) MX2015 (97.9%) PA2013 (80.8%)
PY2016 (92.4%) SR2006 (66.1%) SR2010 (54.0%) SR2018 (77.9%) SV2014 (95.7%)
TT2011 (91.2%) UY2012 (99.7%)
(b) Sani a ion access (dummy)
BZ2006 (92.1%) BZ2011 (95.4%) BZ2015 (92.1%) CR2011 (98.2%) CU2019 (89.7%)
DO2014 (94.9%) DO2019 (92.6%) GY2006 (91.8%) GY2014 (91.2%) GY2019 (95.6%)
HN2019 (85.2%) JM2011 (99.1%) LC2012 (95.9%) MX2015 (97.8%) PA2013 (76.9%)
PY2016 (80.5%) SR2006 (80.0%) SR2010 (69.4%) SR2018 (92.2%) SV2014 (96.2%)
TT2011 (98.3%) UY2012 (98.8%)
No es: This igu e shows he pe cen age o child en who ha e access o wa e and sani a ion se ices in hei homes ac oss di e en coun ies in La in Ame ica
and he Ca ibbean (LAC). Da a we e ob ained om he Mul iple Indica o s Clus e Su ey (MICS) o Belize (BZ), Cos a Rica (CR), Cuba (CU), Dominican
Republic (DO), El Sal ado (SV), Guyana (GY), Hondu as (HN), Jamaica (JM), Mexico (MX), Panama (PA), Pa aguay (PY), Sain Lucia (LC), Su iname (SR),
T inidad and Tobago (TT), and U uguay (UY). We p esen s a is ics o wa e access in Figu e 4a, and sani a ion access in Figu e 4b. S a is ics co espond o
simple na ional a e ages om each su ey.
11
Figu e 5: De elopmen al and heal h ou comes by su ey and access o bo h wa e and sani a ion
(a) ECDI
86.52
83.73
76.19
90.84
81.54
82.32
75.85
81.19
74.57
71.35
90.67
74.40
69.95
77.32
87.50
59.60
70.95
96.30
81.82
89.13
85.13
79.55
93.87
87.94
87.45
82.04
88.86
88.85
74.77
87.82
81.75
82.52
84.15
90.35
73.60
82.06
92.36
89.34
BZ2011
BZ2015
CR2011
CU2019
DO2014
DO2019
GY2014
GY2019
HN2019
JM2011
LC2012
MX2015
PA2013
PY2016
SR2010
SR2018
SV2014
TT2011
UY2012
60 70 80 90 100
No access Access
(b) S un ed
22.12
18.51
22.36
8.12
12.35
24.09
23.82
15.95
12.87
28.62
26.49
12.23
0.00
12.84
11.02
7.53
7.38
19.30
19.60
13.05
6.76
6.52
12.21
12.54
8.61
10.40
17.35
12.28
4.55
2.95
7.12
5.94
4.96
9.85
BZ2006
BZ2011
BZ2015
CU2019
DO2019
GY2006
GY2014
GY2019
HN2019
LC2012
MX2015
PY2016
SR2006
SR2010
SR2018
SV2014
TT2011
0 10 20 30
No access Access
(c) Unde weigh
1.33
3.08
0.83
2.37
3.06
2.19
4.38
5.87
4.78
1.37
1.79
0.66
0.00
3.33
2.91
4.90
4.92
1.58
3.02
1.90
2.47
2.19
2.03
6.03
5.47
5.83
1.87
1.04
0.93
3.32
4.65
4.97
4.84
5.95
BZ2006
BZ2011
BZ2015
CU2019
DO2019
GY2006
GY2014
GY2019
HN2019
LC2012
MX2015
PY2016
SR2006
SR2010
SR2018
SV2014
TT2011
0 2 4 6
No access Access
No es: This cha displays he pe cen age o child en who a e conside ed de elopmen ally on ack acco ding o he ea ly childhood de elopmen index (ECDI),
as well as he a es o s un ing and unde weigh . Figu es o each o hese ca ego ies a e shown in Figu e 5a, Figu e 5b, and Figu e 5c, espec i ely. S a is ics a e
unweigh ed a e ages calcula ed sepa a ely o each su ey o child en wi h and wi hou access o in as uc u e se ices. Da a we e ob ained om he Mul iple
Indica o s Clus e Su ey (MICS) o Belize (BZ), Cos a Rica (CR), Cuba (CU), Dominican Republic (DO), El Sal ado (SV), Guyana (GY), Hondu as (HN),
Jamaica (JM), Mexico (MX), Panama (PA), Pa aguay (PY), Sain Lucia (LC), Su iname (SR), T inidad and Tobago (TT), and U uguay (UY).
12
sani a ion access, wi h a highe pe cen age o child en wi h access o sani a ion being on ack
wi h ECDI in mos coun ies. In nea ly all coun ies, he di e ence in he pe cen age o child en
who a e de elopmen ally on ack is signi ican be ween hose who do and do no ha e access.
The da a also show in Panel B ha he pe cen age o child en wi h s un ing is highe among
hose wi hou access o in as uc u e in mos coun ies, excep T inidad and Tobago and Sain
Lucia. The la ges absolu e di e ence is ound in Hondu as, Mexico and El Sal ado , be ween
hose wi h and wi hou access, while Guyana in 2019 has he smalles di e ence. In Panel C,
we p o ide desc ip i e in o ma ion abou unde weigh , which displays g ea e a ia ion be ween
coun ies compa ed o ECDI and s un ing. Gene ally, child en wi h access o bo h se ices epo
highe incidence o unde weigh han hose wi hou access. Howe e , he e is a ema kably low
incidence o unde weigh ac oss all coun ies in gene al.
O e all, he esul s p esen ed in Figu e 5indica e ha access o wa e and sani a ion se ices
migh be linked o ea ly childhood de elopmen ou comes, wi h child en who ha e access o hese
se ices gene ally ha ing be e ECDI sco es and expe iencing lowe a es o s un ing. We y o
explo e he pa icula low incidence o unde weigh ac oss coun ies du ing he empi ical models.
We employ da a om he Demog aphic and Heal h Su eys (DHS) o examine households’
ga bage managemen in LAC. DHS su eys a e na ionally ep esen a i e household su eys ha
collec da a on a ious aspec s o heal h, including ep oduc i e heal h, child heal h, and nu i ion.
They co e bo h low- and middle-income coun ies, enabling compa a i e analyses ac oss di e en
con ex s. Mo eo e , hese su eys p o ide de ailed in o ma ion on ga bage collec ion, such as ype
and equency, which is ele an o ou hypo hesis ha child en’s heal h ou comes a e nega i ely
a ec ed by highe exposu e o esiden ial was e.
As shown in Figu e 6, we e ie ed ga bage collec ion da a om 14 su eys wi h a o al o 118,597
mo he -child dyads. Access o ga bage collec ion se ices a ies signi ican ly ac oss coun ies. In
Guyana, o ins ance, access o ga bage collec ion was low a 12.1% in 1998, ising o 21% in
2005. Meanwhile, Nica agua epo ed a low access a e o a ound 23% in bo h 1998 and 2001.
The Dominican Republic showed p og ess, wi h he pe cen age o access inc easing om 60.7%
in 2007 o 74.6% in 2013. No ably, access o ga bage collec ion se ices is s eady in Pe u ac oss
mul iple su eys om 1991 o 2011. E en o e wo decades, he pe cen age o access o ga bage
collec ion se ices emained almos s agnan , ho e ing a ound 34%. Addi ional s a is ics o he
pooled sample o coun ies/su eys a e a ailable in Appendix Table A5.
13
Figu e 6: Access o ga bage collec ion by su ey
BO2003 (43.5%) CO2005 (71.4%) CO2010 (55.6%) DR2007 (60.7%) DR2013 (74.6%) GU1998 (12.1%) GU2005 (21.0%)
NC1998 (23.4%) NC2001 (23.6%) PE1991 (32.6%) PE2004 (34.4%) PE2007 (34.4%) PE2009 (34.2%) PE2011 (34.0%)
No es: This igu e p esen s he pe cen age o child en whose households had access o ga bage collec ion se ices
ac oss di e en su eys. Da a we e ob ained om he Demog aphic and Heal h Su eys (DHS) o Boli ia (BO),
Colombia (CO), Dominican Republic (DR), Guyana (GU), Nica agua (NC), and Pe u (PE). These da a only include
coun ies o which ga bage collec ion in o ma ion is a ailable om he DHS su eys. S a is ics ep esen unweigh ed
a e ages.
Figu e 7p esen s di e ences in s un ing and unde weigh by he a ailabili y o ga bage collec ion
ac oss su eys. The la ges di e ence in s un ing be ween child en wi h and wi hou ga bage
collec ion se ices is obse ed in Guyana (1998), a 30.98% and 62.72%, espec i ely, while he
smalles di e ence in s un ing was obse ed in he Dominican Republic (2013), a 7.49% and 10.1%
espec i ely.
Simila ly, child en wi h access o ga bage collec ion se ices a e less equen ly unde weigh han
hose wi hou access. The la ges di e ence in unde weigh child en be ween hose wi h and wi hou
ga bage collec ion se ices was obse ed in Guyana (1998), a 9.36% and 21.25% espec i ely. The
smalles di e ence in unde weigh child en was obse ed in he Dominican Republic (2013), a
3.14% and 4.8% espec i ely. In gene al, hese s a is ics suppo ou hypo hesis ha child en’s
heal h ou comes may be nega i ely a ec ed by highe exposu e o esiden ial was e, highligh ing
he impo ance o in es ing in e icien was e managemen sys ems o p omo ing be e heal h and
ea ly childhood de elopmen ou comes.
4 Resul s
In his sec ion, we p o ide e idence o he po en ially signi ican ole ha access o basic in as uc-
u e se ices, such as wa e , sani a ion, and ga bage collec ion, plays in ea ly childhood heal h and
de elopmen . Speci ically, he MICS da a demons a e a s ong link be ween access o wa e and
sani a ion and child en’s likelihood o being on ack in hei de elopmen acco ding o he ECDI,
as well as a lowe likelihood o s un ing. In addi ion, he DHS da a highligh a nega i e associa-
ion be ween access o imp o ed ga bage collec ion se ices and a es o s un ing and unde weigh
among child en unde i e.
14
Figu e 7: S un ing and unde weigh by ype o ga bage disposal sys em
(a) S un ing
41.48
23.98
26.82
14.55
10.10
62.72
29.30
37.20
31.90
47.91
37.09
37.09
34.58
29.72
23.89
12.59
11.50
9.64
7.49
30.98
16.15
20.83
11.94
19.76
13.84
13.84
13.36
10.92
BO2003
CO2005
CO2010
DR2007
DR2013
GU1998
GU2005
NC1998
NC2001
PE1991
PE2004
PE2007
PE2009
PE2011
10 20 30 40 50 60
S un ing (%)
Non-collec ed
Collec ed
(b) Unde weigh
7.44
7.52
5.78
4.49
4.80
21.25
10.14
11.54
8.58
12.21
5.80
5.80
6.44
6.32
3.36
4.30
2.69
3.15
3.14
9.36
8.50
6.16
3.81
3.44
2.33
2.33
2.51
1.97
BO2003
CO2005
CO2010
DR2007
DR2013
GU1998
GU2005
NC1998
NC2001
PE1991
PE2004
PE2007
PE2009
PE2011
0 5 10 15 20
Unde weigh (%)
Non-collec ed
Collec ed
No es: This igu e p esen s he pe cen age o child en wi h s un ing and unde weigh ac oss di e en su eys by
access o ga bage collec ion. Da a we e ob ained om he Demog aphic and Heal h Su eys (DHS) o Boli ia (BO),
Colombia (CO), Dominican Republic (DR), Guyana (GU), Nica agua (NC), and Pe u (PE). The da a p esen ed he e
only include coun ies o which ga bage collec ion in o ma ion is a ailable om he DHS su eys. S a is ics ep esen
unweigh ed a e ages.
15
Wa e and sani a ion in as uc u e using he MICS: We p esen he esul s o he linea
p obabili y models ha explo e he associa ion be ween access o wa e and sani a ion and a ious
de elopmen al and heal h ou comes. These ou comes include bina y indica o s o being on ack
in he ECDI, as well as s un ing and unde weigh s a us in child en. The da a u ilized o his
analysis o igina e om 22 su eys conduc ed ac oss 14 di e en coun ies. I is impo an o no e
ha he speci ic coun ies and su eys included in he analysis a y based on he model and heal h
ou come in ques ion, as no all ou comes a e a ailable ac oss all su eys.11
Table 1: Access o in as uc u e and he ela ionship wi h he ECDI
Dependen a iable ECDI (dummy)
(1) (2) (3) (4) (5) (6) (7) (8)
Wa e access (dummy) 0.039*** 0.041*** 0.027** 0.019*
(0.010) (0.010) (0.010) (0.010)
Sani a ion access (dummy) 0.059*** 0.058*** 0.037*** 0.029***
(0.009) (0.008) (0.008) (0.008)
R2-squa ed 0.039 0.051 0.059 0.063 0.040 0.052 0.060 0.063
Mean o ou come 0.834 0.834 0.834 0.834 0.834 0.834 0.834 0.834
N 37964 37964 37964 37964 37964 37964 37964 37964
Su eys 19 19 19 19 19 19 19 19
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes Yes Yes
Child en le el No Yes Yes Yes No Yes Yes Yes
Mo he le el No No Yes Yes No No Yes Yes
Household cha ac e is ics No No No Yes No No No Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween child de elopmen
indica o s and access o wa e and sani a ion in as uc u e. The analysis sample is composed om 19 su eys
included in The Mul iple Indica o Clus e Su eys (MICS). The coun ies ha ha e been included in his analysis
and hei espec i e sample sizes can be ound in Appendix Table A1. The analysis con ols o ixed e ec s a
he coun y- egion-yea le el. Addi ional con ols a he child le el include age (in mon hs), bi h coho , and sex.
Mo he -le el con ols accoun o he mo he ’s educa ion le el. Household cha ac e is ics con olled o a e u ban
esidency s a us, access o elec ici y, o al household size, and he numbe o household membe s who a e unde
i e yea s o age. Va iable de ini ions a e a ailable in Appendix Table A2 and summa y s a is ics in Appendix
Table A3. Su ey clus e ed s anda d e o s a e epo ed in pa en heses (*** p<0.01, ** p<0.05, * p<0.1).
In Table 1, we analyze he associa ion be ween access o wa e and sani a ion, and a child’s
de elopmen al p og ess as cap u ed by he ECDI. The models, spanning columns 1 o 8, e eal a
consis en posi i e associa ion. Fo ins ance, he model in column 1, accoun ing o coun y- egion-
yea ixed e ec s, sugges s ha access o wa e co ela es wi h a 3.9% inc ease in he likelihood o a
child being de elopmen ally on ack acco ding o he ECDI. This ela ionship pe sis s e en when
addi ional con ol a iables a e in oduced, including child, mo he , and household cha ac e is ics.
The coe icien s o wa e and sani a ion access sligh ly diminish bu emain signi ican ly posi i e,
11The sample ac oss he di e en models and ou comes is no ha monized due o subs an ial a ia ions in he
numbe and ype o obse a ions when compa ing Tables 1and 2. Despi e his, he esul s emain consis en and
obus when using he same sample, as e idenced by he indings p esen ed in Appendix Table A6.
16
emphasizing he obus ness o hese indings. Child en wi h wa e access exhibi a 1.9% o 4.1%
highe likelihood o op imal de elopmen , while sani a ion access co esponds o a 2.9% o 5.8%
inc ease.
In Appendix Table A7, we e ine his analysis, dissec ing he ECDI in o i s cons i uen sub-
domains: lea ning, socio-emo ional de elopmen , physical de elopmen , and li e acy/nume acy.
The models e eal ha wa e access p edominan ly bene i s lea ning, socio-emo ional de elopmen ,
and li e acy/nume acy, wi h associa ions as high as 7.6%, 1.5%, and 1.9% espec i ely. Sani a ion
access echoes hese bene i s, pa icula ly enhancing lea ning (9.0%), socio-emo ional de elopmen
(2.0%), and li e acy/nume acy (3.4%). The physical de elopmen sub-domain, while showing a
posi i e end, does no each s a is ical signi icance, possibly due o sho - e m measu emen
limi a ions o a iabili y in physical de elopmen miles ones.
In Table 2, we explo e he associa ions be ween access o wa e and sani a ion and heal h ou -
comes, pa icula ly s un ing and being unde weigh in child en. The analysis e eals a no ewo hy
co ela ion: access o wa e is associa ed wi h a 1.3% o 3.8% educ ion in he likelihood o s un ing
and a 0.3% dec ease in he p obabili y o a child being unde weigh . These associa ions a e u he
s eng hened wi h access o sani a ion, which is co ela ed wi h a 4.8% o 8.1% dec ease in he
likelihood o s un ing and a 0.3% educ ion in being unde weigh . Al hough he pe cen age changes
migh seem modes , hey a e signi ican when conside ed in ela ion o he baseline p e alence o
hese heal h ou comes, hus highligh ing hei po en ial ele ance o public heal h e o s.
We es ed addi ional models ha include weal h o mi iga e endogenei y issues and ob ain mo e
accu a e es ima es. Speci ically, weal h may p edic bo h access o imp o ed sani a ion acili ies as
well as de elopmen al and heal h ou comes, leading o issues o selec ion o omi ed a iable bias.
By con olling o weal h, we can accoun o he possibili y ha weal hie indi iduals a e mo e
likely o ha e be e sani a ion acili ies while also simul aneously enjoying be e heal h ou comes
due o o he ac o s ha a e co ela ed wi h weal h. These esul s a e shown in Appendix Table A8.
In ou baseline model, we ha e op ed no o include weal h due o he cons uc ion o he weal h
index, which in eg a es p oxies ela ed o in as uc u e access. This in eg a ion could po en ially
obscu e he independen e ec s o wa e and sani a ion access.12
The es ima es in Appendix Table A8 show ha e en a e con olling o household weal h, wa e
and sani a ion access a e bo h posi i ely associa ed wi h on- ack ECDI sco es and nega i ely
associa ed wi h s un ing and unde weigh . Howe e , only sani a ion is s a is ically signi ican . We
con i m ha he coe icien s o he weal h quin ile dummy a iables show ha child en om iche
households a e mo e likely o be de elopmen ally on ack and ha e lowe a es o s un ing and
unde weigh han hose om poo e households. P e iously, all ou models con ol o ma e nal
educa ion in o de o sepa a e ou he impac o ma e nal educa ion on child de elopmen al and
heal h ou comes. Indeed, ma e nal educa ion is linked o be e hygiene, nu i ion, and heal hca e-
12See Appendix Table A2 o addi ional de ails abou he cons uc ion o he weal h index.
17
Table 2: Access o in as uc u e and he ela ionship wi h heal h ou comes
Dependen a iable S un ed (dummy) Unde weigh (dummy)
(1) (2) (3) (4) (5) (6) (7) (8)
Wa e access (dummy) -0.038***-0.038*** -0.026** -0.013 -0.003 -0.003 -0.003 -0.003
(0.012) (0.012) (0.010) (0.008) (0.003) (0.003) (0.003) (0.003)
R2-squa ed 0.044 0.050 0.060 0.068 0.014 0.017 0.017 0.017
Mean o ou come 0.116 0.116 0.116 0.116 0.027 0.027 0.027 0.027
N 67248 67248 67248 67248 67248 67248 67248 67248
Su eys 17 17 17 17 17 17 17 17
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes Yes Yes
Child en le el No Yes Yes Yes No Yes Yes Yes
Mo he le el No No Yes Yes No No Yes Yes
Household cha ac e is ics No No No Yes No No No Yes
(9) (10) (11) (12) (13) (14) (15) (16)
Sani a ion access (dummy) -0.081***-0.081***-0.061***-0.048*** -0.003 -0.003 -0.002 -0.002
(0.011) (0.011) (0.010) (0.007) (0.002) (0.002) (0.002) (0.002)
R2-squa ed 0.047 0.053 0.062 0.069 0.014 0.017 0.017 0.017
Mean o ou come 0.116 0.116 0.116 0.116 0.027 0.027 0.027 0.027
N 67248 67248 67248 67248 67248 67248 67248 67248
Coun ies 17 17 17 17 17 17 17 17
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes Yes Yes
Child en le el No Yes Yes Yes No Yes Yes Yes
Mo he le el No No Yes Yes No No Yes Yes
Household cha ac e is ics No No No Yes No No No Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween child de elopmen
indica o s and access o wa e and sani a ion in as uc u e. The analysis sample is composed om he su eys
included in he Mul iple Indica o Clus e Su eys (MICS). The coun ies ha ha e been included in his analysis
and hei espec i e sample sizes can be ound in Appendix Table A1. The analysis con ols o ixed e ec s a
he coun y- egion-yea le el. Addi ional con ols a he child le el include age (in mon hs), bi h coho , and sex.
Mo he -le el con ols accoun o he mo he ’s educa ion le el. Household cha ac e is ics con olled o a e u ban
esidency s a us, access o elec ici y, o al household size, and he numbe o household membe s who a e unde
i e yea s o age. Va iable de ini ions a e a ailable in Appendix Table A2 and summa y s a is ics in Appendix
Table A4. Su ey clus e ed s anda d e o s a e epo ed in pa en heses (*** p<0.01, ** p<0.05, * p<0.1).
seeking beha io o child en, and may in luence household weal h and in es men in wa e and
sani a ion in as uc u e. Las ly, we also es ed he obus ness o he esul s o he inclusion o he
s imula ion index, which is a s ong p edic o o ECD (Cua as e al.,2023), ob aining consis en
es ima es.
Recognizing he inhe en in e connec edness o sani a ion and wa e in de e mining heal h and
de elopmen al ou comes, i becomes essen ial o in es iga e no jus hei indi idual in luences,
bu also hei combined e ec s. The esul s o hese in e ac ions a e illus a ed in Figu e 8. By
including an in e ac ion e m be ween sani a ion and wa e , we aim o dis inguish whe he hei
join in luence ampli ies o diminishes he ou comes obse ed when hey ac independen ly. The
18
esul s o ECDI indica e a posi i e associa ion wi h bo h wa e and sani a ion access. Child en
wi h access o bo h wa e and sani a ion ha e a p edic ed 83.8% likelihood o being on ack
in ECDI, which is highe compa ed o hose wi h access o only wa e (81.7%), only sani a ion
(82.7%), o nei he (77.7%). When i comes o s un ing child en wi h access o bo h wa e and
sani a ion ha e he lowes likelihood o being s un ed, a 11.00%. Those wi hou access o ei he
ha e a highe likelihood o 16.75%, Simila ends a e obse ed o child en wi h access o only
wa e (15.76%) o only sani a ion (11.95%). The pa e n con inues o unde weigh , an indica o
o acu e malnu i ion. Child en wi h access o bo h wa e and sani a ion ha e a likelihood o 2.69%,
he lowes among he g oups. Those wi hou access o ei he a e a a highe likelihood o 3.56%.
Access o wa e only esul s in a likelihood o 2.62%, while sani a ion only shows a likelihood o
2.77%.
Figu e 8: P edic ed de elopmen al and heal h ou comes by in as uc u e access
No access o
wa e o
sani a ion
Sani a ion
access only
Wa e access
only
Access o bo h
wa e and
sani a ion
0.74 0.76 0.78 0.80 0.82 0.84 0.10 0.12 0.14 0.16 0.18 0.02 0.03 0.04
ECDI (dummy) S un ed (dummy) Unde weigh (dummy)
P opo ion
No es: The coe icien plo isually displays he p edic i e ma gins o h ee dis inc heal h and de elopmen ou comes
in child en: being on ack acco ding o he ea ly childhood de elopmen index (ECDI), being s un ed, and being
unde weigh . The igu e ca ego izes he esul s based on he child en’s access o wa e and sani a ion, delinea ing
ou scena ios: “No access o wa e o sani a ion”, “Wa e access only”, “Sani a ion access only”, and “Access o
bo h wa e and sani a ion”. The p edic i e ma gins a e calcula ed om he linea p obabili y models in Appendix
Table A9. Va iable de ini ions a e a ailable in Appendix Table A2 and summa y s a is ics in Appendix Tables A3
and A4. Con idence in e als a e ep esen ed by capped lines and se a 90%.
In o de o mi iga e he nega i e impac s o inadequa e access o clean wa e and sani a ion
acili ies, households may adop a ious beha io s o a i udes ha p omo e heal h and well-being.
These p ac ices, anging om ea ing wa e o enhance i s sa e y o consump ion, o ensu ing
he a ailabili y o soap o cleanse s o handwashing, play a c ucial ole in sa egua ding heal h,
pa icula ly in con ex s whe e access o wa e and sani a ion in as uc u e is limi ed. As shown in
Figu e 9a, wa e ea men eme ges as a signi ican mi iga ing ac o , especially o de elopmen al
ou comes. P omo ing household wa e ea men p ac ices could be a iable in e im solu ion while
19
wo king owa ds uni e sal access o wa e and sani a ion. The da a also sugges ha he a ailabili y
o a cleanse migh play a mi iga ing ole, pa icula ly in ela ion o ECDI and s un ing (see Figu e
9b).
The obus ness o ou esul s is ein o ced h ough a se ies o addi ional analyses. Fi s , we
con i m ou esul s h ough he applica ion o a p e e en ial wi hin p opensi y sco e ma ching ap-
p oach (A pino and Cannas,2016), as demons a ed in Appendix Table A13, p o iding u he
con idence in he alidi y and eliabili y o ou indings. Second, we ex end u he ou analyses
wi h he esul s p esen ed in Tables A11 and A12. By employing a ca ego ical a iable o bo h
wa e and sani a ion access, a he han elying on bina y indica o s, we explo e he co ela ion
o di e en ypes o access. Ou indings in e es ingly highligh he signi ican ole o packaged
wa e in mi iga ing he ad e se e ec s associa ed wi h o he o ms o wa e access. This obse a-
ion could be indica i e o he un eliable quali y o piped wa e sou ces in some La in Ame ican
coun ies and he lack o app op ia e s o age op ions, unde sco ing he po en ial need o al e na-
i e sa e d inking wa e solu ions. On he sani a ion on , lush sys ems s and ou as pa icula ly
bene icial, showcasing g ea e e ec i eness in p omo ing heal h and de elopmen al ou comes when
compa ed o bo h hygienic non- lush me hods and unsani a y sys ems, as well as in compa ison
o a comple e lack o sani a ion acili ies. These indings unde line he necessi y o ecognize and
add ess he a ying impac s o di e en ypes o wa e and sani a ion access. They also emphasize
he impo ance o conside ing he quali y and ype o wa e and sani a ion se ices alongside hei
a ailabili y.
The da a om he MICS compellingly illus a e he posi i e ela ionship ha in es men s in
wa e and sani a ion in as uc u e can ha e on he de elopmen and heal h o young child en.
This ela ionship holds s ong e en when aking in o accoun a ious indi idual, ma e nal, and
household ac o s, alongside conside a ions o household weal h and s imula ion ac i i ies. To ge
mo e in o his associa ion, we expand he da a analyses o include in o ma ion on esiden ial was e
collec ion sys ems. Such da a would p o ide a mo e comp ehensi e unde s anding o how was e
managemen in as uc u e se ices impac ea ly childhood de elopmen . Since exposu e o was e
can a ec he quali y o an indi idual’s en i onmen , his addi ional in o ma ion may enhance he
accu acy o he ela ionship be ween in as uc u e se ices and ea ly childhood de elopmen .
Residen ial was e managemen in as uc u e using he DHS: Table 3p esen s he esul s
o he linea p obabili y model o assess he co ela ion be ween ga bage collec ion and child heal h
indica o s, speci ically s un ing and unde weigh . Ou analysis begins wi h a simple co ela ion
using yea -coun y- egion ixed e ec s, which is hen ex ended by in oducing addi ional con ols,
including child en, mo he and household cha ac e is ics. Con olling o household cha ac e is ics
allows us o accoun o he po en ial impac o socio-economic s a us on he ela ionship be ween
ga bage collec ion and heal h indica o s, as discussed p e iously.
20
Figu e 14: P edic ed s un ing and unde weigh o ecen ly sick child en by ype o ga bage collec-
ion
Dia hea=0Dia hea=1
Non-collec ed
Collec ed
Non-collec ed
Collec ed
0.22 0.23 0.24 0.25 0.26 0.27
Cough=0Cough=1
Non-collec ed
Collec ed
Non-collec ed
Collec ed
0.22 0.23 0.24 0.25 0.26 0.27
Fe e =0Fe e =1
Non-collec ed
Collec ed
Non-collec ed
Collec ed
0.22 0.23 0.24 0.25 0.26 0.27
S un ing
Dia hea=0Dia hea=1
Non-collec ed
Collec ed
Non-collec ed
Collec ed
0.05 0.06 0.07 0.08
Cough=0Cough=1
Non-collec ed
Collec ed
Non-collec ed
Collec ed
0.05 0.06 0.07 0.08
Fe e =0Fe e =1
Non-collec ed
Collec ed
Non-collec ed
Collec ed
0.05 0.06 0.07 0.08
Unde weigh
Sick ecen ly × Ga bage collec ion
P opo ion
No es: The coe icien plo isually displays he p edic i e ma gins o wo dis inc heal h ou comes in child en: being
s un ed, and being unde weigh . The esul s a e o ganized based on ecen illness episodes such as dia hea, cough,
and e e in child en. The igu e hen p esen s es ima ions acco ding o he child en’s access o ga bage collec ion. The
p edic i e ma gins a e calcula ed om he linea p obabili y models in Appendix Table A20. Va iable de ini ions a e
a ailable in Appendix Table A2 and summa y s a is ics in Appendix Table A5. Con idence in e als a e ep esen ed
by capped lines and se a 90%.
ligh o hese insigh s, a ge ed policy in e en ions ha add ess was e managemen and sani a ion
in as uc u e, wi h a pa icula ocus on households wi h child en who ha e expe ienced ecen
episodes o illness may be e ec i e. This app oach will help mi iga e he nega i e impac s o poo
was e managemen on s un ing as well as unde weigh s a us, he eby p omo ing imp o ed child
heal h and de elopmen ou comes.
5 Conclusion
This pape p o ides insigh s in o he link be ween access o essen ial in as uc u e se ices and
ea ly childhood de elopmen (ECD) in La in Ame ica and he Ca ibbean (LAC). Ou indings show
ha access o wa e , sani a ion, and ga bage collec ion can signi ican ly impac child en’s de el-
opmen al and heal h ou comes. Speci ically, we documen ha access o wa e and sani a ion is
s ongly associa ed wi h a highe likelihood o being de elopmen ally on ack as measu ed by he
27
ea ly childhood de elopmen index, and wi h ewe ins ances o s un ing and unde weigh . Addi-
ionally, he esul s e eal a nega i e associa ion be ween imp o ed ga bage collec ion se ices and
a es o s un ing and unde weigh among child en unde i e. Ou indings hold s ong e en when
aking in o accoun a ious indi idual, ma e nal, and household ac o s, alongside conside a ions
o household weal h and s imula ion ac i i ies.
Ou indings a e consis en wi h p io e idence (Pipe e al.,2017;Wa e Aid,2016;Schady,
2015;Gao e al.,2021), and ha e impo an implica ions o policymake s in he egion. Fi s and
o emos , he expansion o piped wa e and sani a ion se ices is c ucial o imp o ing child ou -
comes. This expansion aligns wi h e idence indica ing ha imp o emen s in wa e in as uc u e
can subs an ially educe child mo ali y, wi h g ea e e ec s among socioeconomically disad an-
aged g oups. I is essen ial, howe e , o ecognize he complexi y o sepa a ing he e ec s o
imp o ed wa e om ha o imp o ed sani a ion, as he magni ude o he ela ionship o clean
wa e wi h child ou comes is o en di e en han ha o imp o ed sani a ion.
A key dimension o ou analysis ha me i s signi ican emphasis is he di e en ial impac o ypes
o access o in as uc u e, a c ucial aspec conside ing he Sus ainable De elopmen Goals (SDGs)
ha e ele a ed s anda ds o wha cons i u es adequa e access o wa e and sani a ion. Ou analysis
unde sco es he pi o al ole o al e na i es o piped wa e in ei he mi iga ing o exace ba ing
he e ec s o wa e quali y ecei ed by households. This is especially pe inen in con ex s whe e
piped wa e sou ces may be un eliable, highligh ing he need o al e na i e sa e d inking wa e
solu ions. Rega ding sani a ion, lush sys ems demons a e subs an ial bene i s, p omo ing heal h
and de elopmen al ou comes mo e e ec i ely han bo h hygienic non- lush me hods and unsani a y
sys ems, e en in he absence o sani a ion acili ies. These indings s ess he need o ecognize
and add ess he a ying impac s o di e en ypes o wa e and sani a ion access, emphasizing he
necessi y o speci ic measu emen s o quali y and access o essen ial in as uc u e. The e exis
challenges in accu a ely measu ing eal household access, a c i ical conside a ion o policy and
in as uc u e de elopmen .
We unde sco e he impo ance o in es ing in was e managemen in as uc u e as a c i ical bu
o en o e looked componen o comp ehensi e s a egies o enhance ea ly childhood de elopmen
in low- and middle-income coun ies. This is pa icula ly c i ical in ligh o p e ious wo k (Lu
e al.,2020;McCoy e al.,2022) and ou desc ip i e obse a ions o pe sis en inequi ies in access
o basic wa e , sani a ion, and was e disposal se ices in LAC. By ensu ing ha all households
ha e access o all basic se ices ha p omo e good heal h and ounda ional skill de elopmen ,
policymake s can imp o e he de elopmen al ajec o ies and well-being o child en ac oss he
egion and maximize he bene icial impac o o he in es men s (including ea ly childhood ca e and
educa ion and pa en ing p og ams) while p omo ing sus ainable economic de elopmen o yea s
o come.
28
Ou pape also highligh s he need o a ge ed policy in e en ions aimed a imp o ing household
p ac ices in he egion. The li e a u e ocuses on in as uc u e ha dwa e a he han on in e en-
ions ha modi y household beha io s and mi iga ion measu es (Schady,2015;Rupasinghe e al.,
2021). Fo example, households could be encou aged o p o ec hei en i onmen om po en ial
pollu an s h ough mi iga ion measu es such as s o ing was e in lidded con aine s o pa icipa -
ing in communi y-based ecycling p og ams. Simila ly, s a egies like handwashing, and ea ing
wa e hold p omise, bu hei e ec i eness o en elies on beha io al changes. Policymake s could
conside implemen ing public educa ion campaigns o aise awa eness o he impo ance o p ope
household p ac ices o heal hy child de elopmen . Pa en ing and ECD p og ams deli e ed h ough
home isi s and communi y mee ings ha e been expanding in LAC and o e a po en ial means o
p o iding in o ma ion and suppo ela ed o hygiene p ac ices and he p o ec ion o child en om
en i onmen al haza ds (Lee and Lopez-Boo,2019).
Impo an ly, ou indings sugges ha households in he egion can eap signi ican economic
bene i s om in es men s in wa e , sani a ion, and was e managemen in as uc u e and hei
downs eam bene i s o child heal h and de elopmen . Indeed, ecen simula ion s udies es ima e
ha s un ing alone causes lea ning and economic losses wi h a o al cos o app oxima ely $176.8
billion pe bi h coho globally and $44.7 billion in La in Ame ica (Fink e al.,2016), while he
cos o no in es ing in c i ical in e en ions o p omo e ECD in LAC coun ies such as Gua emala,
Nica agua, Colombia, Pe u, Ecuado , and Chile anged om 0.05% o 3.6% o hei g oss domes ic
p oduc (Rich e e al.,2017). The cos o inac ion is he e o e subs an ial and has meaning ul
impac s on indi idual and socie al ou comes. By p omo ing ECD and educing a es o s un ing and
unde weigh among young child en, p ac ices o imp o e wa e , sani a ion, and was e managemen
can lead o lowe heal hca e cos s o amilies, maximize he impac o subsequen in e en ions,
and p omo e highe le els o educa ional a ainmen and wo k o ce p oduc i i y o e ime. The e
is also a necessi y o a mul i-sec o app oach. We unde line he need o join ly add ess a ange
o ac o s –en i onmen al, nu i ional, heal h- ela ed– o enhance ea ly childhood de elopmen
e ec i ely (Rupasinghe e al.,2021;Va gas-Ba on e al.,2019;Black e al.,2017).
Looking ahead, much emains o be lea ned abou he complex ela ionship be ween access o
in as uc u e and ea ly childhood de elopmen in LAC. Fu u e esea ch could explo e he speci ic
mechanisms h ough which access o essen ial in as uc u e se ices a ec s child heal h ou comes,
as well as he po en ial ole o cul u al and social ac o s in shaping household p ac ices in he
egion. Addi ionally, u he s udies could explo e he long- e m economic bene i s o in es ing in
in as uc u e o p omo ing heal hy child de elopmen and sus ainable economic g ow h.
29
Re e ences
Abdul edha, M., R. Al Khadda , D. Jo dan, P. Ko , A. Abdul idha, and K. Hashim (2018). Es-
ima ing solid was e gene a ion by hospi ali y indus y du ing majo es i als: A quan i ica ion
model based on mul iple eg ession. Was e Managemen 77, 388–400.
Acu , K. and D. T. Ka ine (2013). G eenhouse gas emissions, was e and ecycling policy. Jou nal
o En i onmen al Economics and Managemen 65(1), 74–86.
Agyei-Mensah, S. and M. O eng-Ababio (2012). Pe cep ions o heal h and en i onmen al impac s o
e-was e managemen in Ghana. In e na ional Jou nal o En i onmen al Heal h Resea ch 22(6),
500–517.
Al-Kha ib, I. A., I. Abu Fkhidah, J. I. Kha ib, and S. Kon ogianni (2016). Implemen a ion o a
mul i- a iable eg ession analysis in he assessmen o he gene a ion a e and composi ion o
hospi al solid was e o he design o a sus ainable managemen sys em in de eloping coun ies.
Was e Managemen & Resea ch 34(3), 225–234.
Al a ez, L. (2019, June). Access o wa e and sani a ion se ices in dispe sed u al se lemen s:
The pa h o uni e saliza ion. Technical epo .
A pino, B. and M. Cannas (2016). P opensi y sco e ma ching wi h clus e ed da a. An appli-
ca ion o he es ima ion o he impac o caesa ean sec ion on he Apga sco e. S a is ics in
Medicine 35(12), 2074–2091.
A anasio, O., S. Ca an, E. Fi zsimons, C. Meghi , and M. Rubio-Codina (2020). Es ima ing he
p oduc ion unc ion o human capi al: Resul s om a andomized con olled ial in Colombia.
Ame ican Economic Re iew 110(1), 48–85.
Black, M. M., S. P. Walke , L. C. Fe nald, C. T. Ande sen, A. M. DiGi olamo, C. Lu, D. C. McCoy,
G. Fink, Y. R. Shawa , J. Shi man, e al. (2017). Ea ly childhood de elopmen coming o age:
Science h ough he li e cou se. The Lance 389(10064), 77–90.
Bo ns ein, M. H., T. Le en ahl, and R. M. Le ne (2015). Handbook o child psychology and
de elopmen al science: Ecological se ings and p ocesses. Wiley.
Cua as, J. (2021). The e ec o ma e nal educa ion on pa en ing and ea ly childhood de elopmen :
An ins umen al a iables app oach. Jou nal o Family Psychology.
Cua as, J. (2022). The e ec o spanking on ea ly social-emo ional skills. Child De elopmen 93(1),
180–193.
Cua as, J., D. McCoy, J. S´anchez, J. Beh man, C. Cappa, G. Dona i, J. Heymann, C. Lu,
A. Raikes, N. Rao, e al. (2023). Family play, eading, and o he s imula ion and ea ly childhood
de elopmen in i e low-and-middle-income coun ies. De elopmen al Science, e13404.
Cua as, J., C. Rey-Gue a, D. C. McCoy, and E. Hanno (2020). Ma e nal knowledge, s imula ion,
and ea ly childhood de elopmen in low-income amilies in Colombia. In ancy 25(5), 526–534.
Cua as, J., D. G. Weissman, M. A. She idan, L. Lengua, and K. A. McLaughlin (2021). Co po al
punishmen and ele a ed neu al esponse o h ea in child en. Child De elopmen 92(3), 821–
832.
Cunha, F. and J. Heckman (2007). The echnology o skill o ma ion. Ame ican Economic Re-
iew 97(2), 31–47.
Danaei, G., K. G. And ews, C. R. Sud eld, G. Fink, D. C. McCoy, E. Pee , A. Sania, M. C.
Smi h Fawzi, M. Ezza i, and W. W. Fawzi (2016). Risk ac o s o childhood s un ing in 137
de eloping coun ies: A compa a i e isk assessmen analysis a global, egional, and coun y
le els. PLoS medicine 13(11), e1002164.
30
Engle, P. L., M. M. Black, J. R. Beh man, M. C. De Mello, P. J. Ge le , L. Kapi i i, R. Ma o ell,
M. E. Young, I. C. D. S. G oup, e al. (2007). S a egies o a oid he loss o de elopmen al
po en ial in mo e han 200 million child en in he de eloping wo ld. The Lance 369(9557),
229–242.
Fink, G., E. Pee , G. Danaei, K. And ews, D. C. McCoy, C. R. Sud eld, M. C. Smi h Fawzi,
M. Ezza i, and W. W. Fawzi (2016). Schooling and wage income losses due o ea ly-childhood
g ow h al e ing in de eloping coun ies: Na ional, egional, and global es ima es. The Ame ican
Jou nal o Clinical Nu i ion 104(1), 104–112.
Gao, Y., L. Zhang, A. Kc, Y. Wang, S. Zou, C. Chen, Y. Huang, X. Mi, and H. Zhou (2021).
Housing en i onmen and ea ly childhood de elopmen in sub-saha an A ica: A c oss-sec ional
analysis. PLoS Medicine 18(4), e1003578.
G au, J., H. Te aza, D. M. R. Velosa, A. Rihm, and G. S u zenegge (2015). Solid was e manage-
men in La in Ame ica and he Ca ibbean.
G¨un he , I. and R. Laho i (2021). T ansi ioning o No Po e y. MDPI-Mul idisciplina y Digi al
Publishing Ins i u e.
Hamil on, S. F., T. W. Sp oul, D. Sunding, and D. Zilbe man (2013). En i onmen al policy wi h
collec i e was e disposal. Jou nal o En i onmen al Economics and Managemen 66(2), 337–346.
Heckman, J. J. (2006). Skill o ma ion and he economics o in es ing in disad an aged child en.
Science 312(5782), 1900–1902.
Heindel, J. J. and T. T. Schug (2014). The obesogen hypo hesis: Cu en s a us and implica ions
o human heal h. Cu en En i onmen al Heal h Repo s 1(4), 333–340.
Janmaimool, P. (2017). Applica ion o p o ec ion mo i a ion heo y o in es iga e sus ainable was e
managemen beha io s. Sus ainabili y 9(7), 1079.
Jeong, J., D. C. McCoy, and G. Fink (2017). Pa hways be ween pa e nal and ma e nal educa-
ion, ca egi e s’ suppo o lea ning, and ea ly child de elopmen in 44 low-and middle-income
coun ies. Ea ly Childhood Resea ch Qua e ly 41, 136–148.
Jeong, J., D. C. McCoy, A. K. Yousa zai, C. Salhi, and G. Fink (2016). Pa e nal s imula ion and
ea ly child de elopmen in low-and middle-income coun ies. Pedia ics 138(4).
Jeuland, M. A., D. E. Fuen e, S. Ozdemi , M. C. Allai e, and D. Whi ing on (2013). The long- e m
dynamics o mo ali y bene i s om imp o ed wa e and sani a ion in less de eloped coun ies.
PloS one 8(10), e74804.
Kese , S., S. Duzgun, and A. Aksoy (2012). Applica ion o spa ial and non-spa ial da a analysis in
de e mina ion o he ac o s ha impac municipal solid was e gene a ion a es in u key. Was e
Managemen 32(3), 359–371.
Knudsen, E. I., J. J. Heckman, J. L. Came on, and J. P. Shonko (2006). Economic, neu obiological,
and beha io al pe spec i es on building ame ica’s u u e wo k o ce. P oceedings o he Na ional
Academy o Sciences 103(27), 10155–10162.
Ko c, M. and F. Hauchman (2021). Ad ancing en i onmen al public heal h in La in Ame ica and
he Ca ibbean. Re is a Paname icana de Salud P´ublica 45.
Lee , J. and F. Lopez-Boo (2019). Assessing he quali y o home isi pa en ing p og ams in La in
Ame ica and he Ca ibbean. Ea ly Child De elopmen and Ca e 189(13), 2183–2196.
Loizillon, A., N. Pe owski, P. B i o, and C. Cappa (2017). De elopmen o he ea ly childhood
de elopmen index in MICS su eys. MICS Me hodological Pape s, No. 6, Da a and Analy ics
Sec ion, Di ision o Da a. Resea ch and Policy, UNICEF New Yo k, 29.
Lu, C., J. Cua as, G. Fink, D. McCoy, K. Liu, Z. Li, B. Daelmans, and L. Rich e (2020).
Inequali ies in ea ly childhood ca e and de elopmen in low/middle-income coun ies: 2010–
2018. BMJ Global Heal h 5(2), e002314.
31
Manga, V. E., O. T. O u, and M. N. Ngwabie (2019). Household pe cep ion and willingness o pay
o imp o ed was e managemen se ice in Mam e, Came oon. A ican Jou nal o En i onmen al
Science and Technology 13(9), 354–364.
Ma el, P. (2016). Re iew o op ions o epo ing wa e , sani a ion and hygiene co e age by weal h
quin ile. New Yo k: UNICEF Da a and Analy ics Sec ion, Di ision o Da a, Resea ch and Policy.
McCoy, D. C., A. L. Roy, and C. C. Ra e (2016). Neighbo hood c ime as a p edic o o indi idual
di e ences in emo ional p ocessing and egula ion. De elopmen al science 19(1), 164–174.
McCoy, D. C., J. Seiden, J. Cua as, L. Pisani, and M. Waldman (2022). Es ima es o a mul-
idimensional index o nu u ing ca e in he nex 1000 days o li e o child en in low-income
and middle-income coun ies: A modelling s udy. The Lance Child & Adolescen Heal h 6(5),
324–334.
Pa il, S. R., B. F. A nold, A. L. Sal a o e, B. B iceno, S. Ganguly, J. M. Col o d J , and P. J.
Ge le (2014). The e ec o India’s o al sani a ion campaign on de eca ion beha io s and child
heal h in u al Madhya P adesh: A clus e andomized con olled ial. PLoS medicine 11(8),
e1001709.
Pe umal, N., D. G. Bassani, and D. E. Ro h (2018). Use and misuse o s un ing as a measu e o
child heal h. The Jou nal o Nu i ion 148(3), 311–315.
Pipe , J. D., J. Chandna, E. Allen, K. Linkman, O. Cumming, A. J. P ende gas , and M. J. Glad-
s one (2017). Wa e , sani a ion and hygiene (wash) in e en ions: E ec s on child de elopmen
in low-and middle-income coun ies. The Coch ane Da abase o Sys ema ic Re iews 2017(3).
Ra hnamala, G., R. Ashwini, and N. Babi ha (2021). Domes ic en i onmen al des uc ions due o
lack o solid was e managemen in u al a eas. Ad . Ma h. Sci. J 10(3), 1807–1819.
Rich e , L. M., B. Daelmans, J. Lomba di, J. Heymann, F. L. Boo, J. R. Beh man, C. Lu, J. E.
Lucas, R. Pe ez-Escamilla, T. Dua, e al. (2017). In es ing in he ounda ion o sus ainable
de elopmen : Pa hways o scale up o ea ly childhood de elopmen . The Lance 389(10064),
103–118.
Rupasinghe, N., L. Ben on, K. Mande ille, and M. La ypo a (2021). Lessons om a e iew o
in e na ional expe ience in ea ly childhood de elopmen p og amming, pa 1.
Salhi, C., E. Bea iz, R. McBain, D. McCoy, M. She idan, and G. Fink (2021). Physical discipline,
dep i a ion, and di e en ial isk o de elopmen al delay ac oss 17 coun ies. Jou nal o he
Ame ican Academy o Child & Adolescen Psychia y 60(2), 296–306.
Sania, A., C. R. Sud eld, G. Danaei, G. Fink, D. C. McCoy, Z. Zhu, M. C. S. Fawzi, M. Akman,
S. E. A i een, A. J. Ba os, e al. (2019). Ea ly li e isk ac o s o mo o , cogni i e and language
de elopmen : A pooled analysis o s udies om low/middle-income coun ies. BMJ open 9(10),
e026449.
Schady, N. (2015). Does access o be e wa e and sani a ion in as uc u e imp o e child ou -
comes? E idence om La in Ame ica and he Ca ibbean. Technical epo , BID.
S¸chiopu, A.-M., I. Apos ol, M. Hodo eanu, and M. Ga ilescu (2007). Solid was e in omania:
Managemen , ea men and pollu ion p e en ion p ac ices. En i onmen al Enginee ing & Man-
agemen Jou nal (EEMJ) 6 (5).
She idan, M. and C. A. Nelson (2009). Neu obiology o e al and in an de elopmen . Handbook o
In an Men al Heal h 3, 40–58.
Spa kman, D. and G. S u zenegge (2016). Fos e ing wa e and sani a ion ma ke s in La in Ame ica
and he Ca ibbean: How he public sec o can suppo he p i a e sec o o b idge co e age gaps
and imp o e se ice quali y o low-income popula ions.
32
S ewa , C. P., P. Ka ige , L. Fe nald, A. J. Picke ing, C. D. A nold, B. F. A nold, A. E. Hubba d,
H. N. Den z, A. Lin, T. J. Mee ke k, e al. (2018). E ec s o wa e quali y, sani a ion, handwash-
ing, and nu i ional in e en ions on child de elopmen in u al Kenya (WASH Bene i s Kenya):
A clus e - andomised con olled ial. The lance child & adolescen heal h 2(4), 269–280.
T en acos a, C. J., P. Da is-Kean, C. Mi chell, L. Hyde, and D. Dolinoy (2016). En i onmen al
con aminan s and child de elopmen . Child De elopmen Pe spec i es 10(4), 228–233.
UNICEF e al. (2020). A new ool o measu e sdg indica o 4.2. Ea ly Childhood De elopmen
Index 2030.
Uni ed Na ions, U. (2015). T ans o ming ou wo ld: The 2030 agenda o sus ainable de elopmen .
Vacca i, M., T. Tudo , and G. Vin i (2019). Cha ac e is ics o leacha e om land ills and dumpsi es
in Asia, A ica and La in Ame ica: An o e iew. Was e Managemen 95, 416–431.
Va gas-Ba on, E., J. Small, D. We lieb, H. Hix-Small, R. G´omez Bo e o, K. Diehl, P. Ve ga a, and
P. Lynch (2019). Global su ey o inclusi e ea ly childhood de elopmen and ea ly childhood
in e en ion p og ams. RISE Ins i u e: Washing on, DC, USA.
Walke , S. P., T. D. Wachs, S. G an ham-McG ego , M. M. Black, C. A. Nelson, S. L. Hu man,
H. Bake -Henningham, S. M. Chang, J. D. Hamadani, B. Lozo , e al. (2011). Inequali y in
ea ly childhood: Risk and p o ec i e ac o s o ea ly child de elopmen . The Lance 378(9799),
1325–1338.
Wang, F., Z. Cheng, A. Reisne , and Y. Liu (2018). Compliance wi h household solid was e
managemen in u al illages in de eloping coun ies. Jou nal o Cleane P oduc ion 202, 293–
298.
Wa e Aid (2016). The missing ing edien s: A e policy-make s doing enough on wa e , sani a ion
and hygiene o end malnu i ion? Field Exchange 53, 56.
WHO, W. (2009). WHO child g ow h s anda ds: G ow h eloci y based on weigh , leng h and head
ci cum e ence: Me hods and de elopmen . Wo ld Heal h O ganiza ion.
WHO, W. (2019). Wa e , sani a ion, hygiene and heal h: A p ime o heal h p o essionals.
Wol , S. and D. C. McCoy (2019). Household socioeconomic s a us and pa en al in es men s:
Di ec and indi ec ela ions wi h school eadiness in ghana. Child De elopmen 90(1), 260–278.
Yamamo o, M. and T. C. Kinnaman (2022). Is incine a ion ep essing ecycling? Jou nal o
En i onmen al Economics and Managemen 111, 102593.
33
Appendix
Figu e A1: P edic ed heal h ou comes by weal h quin ile and ype o ga bage collec ion
Poo es Poo e MiddleRiche Riches
Non-collec ed
Collec ed
Non-collec ed
Collec ed
Non-collec ed
Collec ed
Non-collec ed
Collec ed
Non-collec ed
Collec ed
0.10 0.15 0.20 0.25 0.30 0.02 0.04 0.06 0.08
S un ed (dummy) Unde weigh (dummy)
P opo ion
No es: The coe icien plo isually displays he p edic i e ma gins o wo dis inc heal h ou comes in child en:
being s un ed, and being unde weigh . The esul s a e ca ego ized based on he household’s weal h quin ile. The
igu e hen p esen s es ima ions acco ding o he child en’s access o ga bage collec ion. The p edic i e ma gins a e
calcula ed om he linea p obabili y models in Appendix Table A14. Va iable de ini ions a e a ailable in Appendix
Table A2 and summa y s a is ics in Appendix Table A5. Con idence in e als a e ep esen ed by capped lines and
se a 90%.
34
Figu e A2: P edic ed s un ing and unde weigh wi h di e en household in as uc u e access by
ype o ga bage collec ion
O -g idOn-g id
Non-collec ed
Collec ed
Non-collec ed
Collec ed
.22 .24 .26 .28 .3 .32
Elec ici y
PipedNa u eO he
Non-collec ed
Collec ed
Non-collec ed
Collec ed
Non-collec ed
Collec ed
.2 .22 .24 .26 .28
Wa e
Ou sideInside
Non-collec ed
Collec ed
Non-collec ed
Collec ed
.15 .2 .25 .3
Sewage
S un ing
O -g idOn-g id
Non-collec ed
Collec ed
Non-collec ed
Collec ed
.02 .04 .06 .08 .1
PipedNa u eO he
Non-collec ed
Collec ed
Non-collec ed
Collec ed
Non-collec ed
Collec ed
.02 .04 .06 .08 .1
Ou sideInside
Non-collec ed
Collec ed
Non-collec ed
Collec ed
.02 .04 .06 .08 .1
Unde weigh
P opo ion
No es: The coe icien plo isually displays he p edic i e ma gins o wo dis inc heal h ou comes in child en:
being s un ed, and being unde weigh . The esul s a e ca ego ized based on he household’s weal h quin ile. The
igu e hen p esen s es ima ions acco ding o he child en’s access o ga bage collec ion. The p edic i e ma gins a e
calcula ed om he linea p obabili y models in Appendix Table A16. Va iable de ini ions a e a ailable in Appendix
Table A2 and summa y s a is ics in Appendix Table A5. Con idence in e als a e ep esen ed by capped lines and
se a 90%.
35
Table A1: Sample size by da ase and ou come
MICS DHS
ECDI Heal h Heal h
Coun y Yea N Coun y Yea N Coun y Yea N
Belize 2006 796 Belize 2011 785 Boli ia 2003 9,165
2011 1,945 2015 1,090 Colombia 2005 12,480
2015 2,537 Cos a Rica 2011 906 2010 24,560
Cuba 2019 5,252 Cuba 2019 2,316 Dominican Republic 2007 9,221
Dominican Republic 2019 8,396 Dominican Republic 2014 7,781 2013 3,116
El Sal ado 2014 7,337 2019 3,404 Guyana 1998 3,982
Guyana 2006 2,462 El Sal ado 2014 2,986 2005 1,684
2014 3,335 Guyana 2014 1,307 Nica agua 1998 6,935
2019 2,782 2019 1,184 2001 6,008
Hondu as 2019 8,464 Hondu as 2019 3,616 Pe u 1991 7,757
Mexico 2015 8,065 Jamaica 2011 666 2004 7,990
Pa aguay 2016 4,614 Mexico 2015 3,352 2007 7,990
Sain Lucia 2012 291 Panama 2013 2,305 2009 9,199
Su iname 2006 2,248 Pa aguay 2016 1,830 2011 8,510
2010 3,301 Sain Lucia 2012 122
2018 4,225 Su iname 2010 1,265
T inidad and Tobago 2011 1,198 2018 1,772
T inidad and Tobago 2011 525
U uguay 2012 752
No es: This able shows he sample sizes o each coun y, g ouped by ou come and da ase . The da ase used in he
analysis a e he Mul iple Indica o s Clus e Su eys (MICS) and he Demog aphic Heal h Su eys (DHS).
36
Table A4: MICS s a is ics o heal h analysis
Va iable Type N µ σ Min Max
2014 67,248 0.094 0.292 0.000 1.000
2015 67,248 0.130 0.336 0.000 1.000
2016 67,248 0.097 0.296 0.000 1.000
2017 67,248 0.082 0.275 0.000 1.000
2018 67,248 0.079 0.270 0.000 1.000
2019 67,248 0.041 0.199 0.000 1.000
Elec ici y dummy 67,248 0.903 0.296 0.000 1.000
Female dummy 67,248 0.492 0.500 0.000 1.000
Hands cleanse a ailable dummy 59,142 0.931 0.253 0.000 1.000
Household membe s numbe 67,248 5.442 2.348 2.000 32.000
Household membe s ≤5 numbe 67,248 1.499 0.731 1.000 9.000
Mo he ’s educa ion ca ego ical
Less han p ima y 67,248 0.047 0.211 0.000 1.000
P ima y 67,248 0.286 0.452 0.000 1.000
Lowe seconda y 67,248 0.158 0.364 0.000 1.000
Uppe seconda y 67,248 0.071 0.256 0.000 1.000
Seconda y 67,248 0.303 0.459 0.000 1.000
Te ia y/highe /uni e si y 67,248 0.130 0.337 0.000 1.000
Voca ional/ echnical 67,248 0.000 0.013 0.000 1.000
O he o non-s anda d cu iculum 67,248 0.002 0.042 0.000 1.000
S imula ion index ca ego ical 36,760 4.240 1.934 0.000 6.000
U ban a ea dummy 67,248 0.506 0.500 0.000 1.000
Wa e ea ed o sa e d inking dummy 67,171 0.308 0.462 0.000 1.000
Weal h index quin iles ca ego ical
Poo es 67,248 0.313 0.464 0.000 1.000
Second 67,248 0.215 0.411 0.000 1.000
Middle 67,248 0.186 0.389 0.000 1.000
Fou h 67,248 0.161 0.368 0.000 1.000
Riches 67,248 0.126 0.331 0.000 1.000
No es: This able p esen s desc ip i e s a is ics o he a iables om he UNICEF’s Mul iple
Indica o s Clus e Su ey (MICS). The analysis sample is composed om wel e su eys.
In pa icula , A gen ina 2011, Belize 2015, Cos a Rica 2018, Dominican Republic 2014, El
Sal ado 2014, Guyana 2014, Hai i 2016, Jamaica 2011, Mexico 2015, Panama 2013, Pa aguay
2016, and U uguay 2012.
43
Table A5: DHS s a is ics
Va iable Type N µ σ Min Max
Ou comes and was e managemen
S un ing dummy 118,597 0.251 0.434 0.000 1.000
Unde weigh dummy 118,597 0.060 0.237 0.000 1.000
Ga bage collec ion dummy 118,597 0.442 0.497 0.000 1.000
Con ols
Age in mon hs numbe 118,597 29.770 17.216 0.000 59.000
Coho ca ego ical
1986 118,597 0.000 0.021 0.000 1.000
1987 118,597 0.012 0.110 0.000 1.000
1988 118,597 0.014 0.116 0.000 1.000
1989 118,597 0.013 0.112 0.000 1.000
1990 118,597 0.012 0.111 0.000 1.000
1991 118,597 0.014 0.116 0.000 1.000
1992 118,597 0.000 0.020 0.000 1.000
1993 118,597 0.010 0.100 0.000 1.000
1994 118,597 0.017 0.130 0.000 1.000
1995 118,597 0.018 0.134 0.000 1.000
1996 118,597 0.019 0.138 0.000 1.000
1997 118,597 0.029 0.167 0.000 1.000
1998 118,597 0.021 0.143 0.000 1.000
1999 118,597 0.028 0.166 0.000 1.000
2000 118,597 0.043 0.203 0.000 1.000
2001 118,597 0.043 0.204 0.000 1.000
2002 118,597 0.049 0.217 0.000 1.000
2003 118,597 0.068 0.252 0.000 1.000
2004 118,597 0.074 0.263 0.000 1.000
2005 118,597 0.088 0.283 0.000 1.000
2006 118,597 0.108 0.311 0.000 1.000
2007 118,597 0.100 0.300 0.000 1.000
2008 118,597 0.086 0.280 0.000 1.000
2009 118,597 0.072 0.258 0.000 1.000
2010 118,597 0.038 0.192 0.000 1.000
2011 118,597 0.013 0.111 0.000 1.000
2012 118,597 0.006 0.075 0.000 1.000
2013 118,597 0.003 0.058 0.000 1.000
Cough dummy 118,445 0.376 0.484 0.000 1.000
Dia hea dummy 118,450 0.161 0.368 0.000 1.000
Elec ici y dummy 118,597 0.757 0.429 0.000 1.000
Female dummy 118,597 0.492 0.500 0.000 1.000
Fe e dummy 110,652 0.256 0.436 0.000 1.000
Ga bage con aine ca ego ical
Lidded 42,199 0.135 0.342 0.000 1.000
Lidless 42,199 0.359 0.480 0.000 1.000
44
Table A5: DHS s a is ics
Va iable Type N µ σ Min Max
Bags 42,199 0.382 0.486 0.000 1.000
O he 42,199 0.123 0.328 0.000 1.000
Re ige a o 118,597 0.350 0.477 0.000 1.000
Household membe s numbe 118,597 5.962 2.531 2.000 30.000
Household membe s ≤5 numbe 118,597 1.732 0.859 0.000 10.000
Mo he ’s educa ion qua iles ca ego ical
Low 118,597 0.280 0.449 0.000 1.000
In e qua ile ange 118,597 0.561 0.496 0.000 1.000
Top 118,597 0.159 0.366 0.000 1.000
Mo he is head o household dummy 118,597 0.096 0.294 0.000 1.000
Toile is inside dwelling dummy 118,597 0.357 0.479 0.000 1.000
U ban a ea dummy 118,597 0.552 0.497 0.000 1.000
Wa e access ca ego ical
Piped wa e sou ce 118,597 0.616 0.486 0.000 1.000
Na u e sou ces 118,597 0.252 0.434 0.000 1.000
O he 118,597 0.132 0.338 0.000 1.000
No es: This able p esen s desc ip i e s a is ics o he a iables om he Demog aphic
and Heal h Su eys (DHS). The analysis sample is composed om ou een su eys. In
pa icula , Boli ia (2003), Colombia (2005, and 2010), Dominican Republic (2007, and
2013), Guyana (1998, and 2005), Nica agua (1998, and 2001), and Pe u (1991, 2004,
2007, 2009, and 2011). Va iable de ini ions a e a ailable in Appendix Table A2.
45
Table A6: Access o in as uc u e and he ela ionship wi h de elopmen al and heal h ou comes: ha monizing he sample
Dependen a iable ECDI (dummy) S un ed (dummy) Unde weigh (dummy)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Wa e access (dummy) 0.041*** 0.042*** 0.030** 0.023 -0.040** -0.040** -0.028** -0.013 -0.001 -0.001 -0.002 -0.002
(0.013) (0.013) (0.014) (0.013) (0.013) (0.013) (0.011) (0.008) (0.005) (0.005) (0.005) (0.005)
R2-squa ed 0.044 0.057 0.065 0.068 0.066 0.069 0.082 0.089 0.021 0.023 0.023 0.023
Mean o ou come 0.824 0.824 0.824 0.824 0.109 0.109 0.109 0.109 0.023 0.023 0.023 0.023
N 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599
Su eys 14 14 14 14 14 14 14 14 14 14 14 14
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Child en le el No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes
Mo he le el No No Yes Yes No No Yes Yes No No Yes Yes
Household cha ac e is ics No No No Yes No No No Yes No No No Yes
(13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24)
Sani a ion access (dummy) 0.052*** 0.051*** 0.032*** 0.025*** -0.077*** -0.077*** -0.056*** -0.043*** -0.002 -0.002 -0.002 -0.002
(0.008) (0.008) (0.008) (0.008) (0.016) (0.016) (0.013) (0.010) (0.002) (0.002) (0.002) (0.002)
R2-squa ed 0.044 0.058 0.065 0.068 0.069 0.072 0.083 0.091 0.021 0.023 0.023 0.023
Mean o ou come 0.824 0.824 0.824 0.824 0.109 0.109 0.109 0.109 0.023 0.023 0.023 0.023
N 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599 25599
Su eys 14 14 14 14 14 14 14 14 14 14 14 14
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Child en le el No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes
Mo he le el No No Yes Yes No No Yes Yes No No Yes Yes
Household cha ac e is ics No No No Yes No No No Yes No No No Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween he ea ly childhood de elopmen index (ECDI) and access o wa e and
sani a ion in as uc u e. The analysis sample is composed om ou een su eys included in he Mul iple Indica o Clus e Su eys (MICS). The analysis con ols o
ixed e ec s a he coun y- egion-yea le el. Addi ional con ols a he child le el include age (in mon hs), bi h coho , and sex. Mo he -le el con ols accoun o he
mo he ’s educa ion le el. Household cha ac e is ics con olled o a e u ban esidency s a us, access o elec ici y, o al household size, and he numbe o household
membe s who a e unde i e yea s o age. Va iable de ini ions a e a ailable in Appendix Table A2. Su ey clus e ed s anda d e o s a e epo ed in pa en heses (***
p<0.01, ** p<0.05, * p<0.1).
Table A7: Access o in as uc u e and he ela ionship wi h he de elopmen al componen s o he ECDI
Dependen a iable ECDI lea ning (dummy) ECDI physical (dummy) ECDI socio-emo ional (dummy) ECDI li e acy/nume acy (dummy)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
Wa e access (dummy) 0.076*** 0.078*** 0.054*** 0.039** 0.004 0.004 0.003 0.003 0.014* 0.015* 0.009 0.006 0.018** 0.019** 0.013* 0.008
(0.015) (0.015) (0.015) (0.015) (0.003) (0.003) (0.004) (0.003) (0.008) (0.008) (0.009) (0.008) (0.007) (0.007) (0.007) (0.006)
R2-squa ed 0.159 0.197 0.215 0.218 0.008 0.009 0.010 0.011 0.042 0.049 0.052 0.055 0.023 0.026 0.033 0.035
Mean o ou come 0.275 0.275 0.275 0.275 0.985 0.985 0.985 0.985 0.815 0.815 0.815 0.815 0.970 0.970 0.970 0.970
N 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964
Su eys 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Child en le el No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes
Mo he le el No No Yes Yes No No Yes Yes No No Yes Yes No No Yes Yes
Household cha ac e is ics No No No Yes No No No Yes No No No Yes No No No Yes
(17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32)
Sani a ion access (dummy) 0.090*** 0.088*** 0.050*** 0.035*** 0.004 0.004 0.003 0.002 0.020** 0.019** 0.010 0.006 0.034*** 0.034*** 0.025*** 0.020**
(0.009) (0.009) (0.006) (0.007) (0.003) (0.003) (0.003) (0.003) (0.009) (0.009) (0.009) (0.009) (0.007) (0.007) (0.007) (0.007)
R2-squa ed 0.159 0.197 0.215 0.218 0.008 0.009 0.010 0.011 0.042 0.049 0.052 0.055 0.025 0.028 0.033 0.036
Mean o ou come 0.275 0.275 0.275 0.275 0.985 0.985 0.985 0.985 0.815 0.815 0.815 0.815 0.970 0.970 0.970 0.970
N 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964 37964
Coun ies 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Child en le el No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes No Yes Yes Yes
Mo he le el No No Yes Yes No No Yes Yes No No Yes Yes No No Yes Yes
Household cha ac e is ics No No No Yes No No No Yes No No No Yes No No No Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween he sub-componen s o he ea ly childhood de elopmen index (ECDI)
and access o wa e and sani a ion in as uc u e. The analysis sample is composed om 19 su eys included in The Mul iple Indica o Clus e Su eys (MICS). The
coun ies ha ha e been included in his analysis and hei espec i e sample sizes can be ound in Appendix Table A1. The analysis con ols o ixed e ec s a
he coun y- egion-yea le el. Addi ional con ols a he child le el include age (in mon hs), bi h coho , and sex. Mo he -le el con ols accoun o he mo he ’s
educa ion le el. Household cha ac e is ics con olled o a e u ban esidency s a us, access o elec ici y, o al household size, and he numbe o household membe s
who a e unde i e yea s o age. Va iable de ini ions a e a ailable in Appendix Table A2 and summa y s a is ics in Appendix Table A3. Su ey clus e ed s anda d
e o s a e epo ed in pa en heses (*** p<0.01, ** p<0.05, * p<0.1).
Table A8: Access o in as uc u e and he ela ionship wi h de elopmen al and heal h ou comes:
weal h and s imula ion
Dependen a iable ECDI S un ed Unde weigh ECDI S un ed Unde weigh
(dummy) (dummy) (dummy) (dummy) (dummy) (dummy)
(1) (2) (3) (4) (5) (6)
Wa e access (dummy) 0.017 -0.009 -0.001 0.011 -0.002 -0.003
(0.010) (0.007) (0.003) (0.010) (0.008) (0.003)
R2-squa ed 0.067 0.090 0.022 0.065 0.072 0.018
Mean o ou come 0.834 0.117 0.026 0.834 0.116 0.027
N 37951 36760 36760 37964 67248 67248
Coun ies 19 17 17 19 17 17
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes
Child en le el Yes Yes Yes Yes Yes Yes
Mo he le el Yes Yes Yes Yes Yes Yes
Household cha ac e is ics Yes Yes Yes Yes Yes Yes
S imula ion index Yes Yes Yes No No No
Weal h quin iles No No No Yes Yes Yes
(7) (8) (9) (10) (11) (12)
Sani a ion access (dummy) 0.024*** -0.051*** -0.004 0.018* -0.032*** -0.003
(0.007) (0.013) (0.003) (0.009) (0.007) (0.002)
R2-squa ed 0.067 0.092 0.022 0.065 0.073 0.018
Mean o ou come 0.834 0.117 0.026 0.834 0.116 0.027
N 37951 36760 36760 37964 67248 67248
Su eys 19 17 17 19 17 17
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes
Child en le el Yes Yes Yes Yes Yes Yes
Mo he le el Yes Yes Yes Yes Yes Yes
Household cha ac e is ics Yes Yes Yes Yes Yes Yes
S imula ion index Yes Yes Yes No No No
Weal h quin iles No No No Yes Yes Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween child
de elopmen and heal h indica o s and access o wa e and sani a ion in as uc u e including weal h
and s imula ion ac o s as con ols. The analysis sample is composed om mul iple su eys included in
he Mul iple Indica o Clus e Su eys (MICS). The coun ies ha ha e been included in his analysis
and hei espec i e sample sizes can be ound in Appendix Table A1. The analysis con ols o ixed
e ec s a he coun y- egion-yea le el. Addi ional con ols a he child le el include age (in mon hs), bi h
coho , and sex. Mo he -le el con ols accoun o he mo he ’s educa ion le el. Household cha ac e is ics
con olled o a e u ban esidency s a us, access o elec ici y, o al household size, and he numbe o
household membe s who a e unde i e yea s o age. Va iable de ini ions a e a ailable in Appendix Table
A2 and summa y s a is ics in Appendix Tables A3 and A4. Su ey clus e ed s anda d e o s a e epo ed
in pa en heses (*** p<0.01, ** p<0.05, * p<0.1).
48
Table A9: Access o in as uc u e and he ela ionship wi h de elopmen al and heal h ou comes:
in e ac ions be ween wa e and sani a ion
ECDI S un ed Unde weigh
(dummy) (dummy) (dummy)
(1) (2) (3)
Wa e access (dummy) 0.040 -0.010 -0.009*
(0.026) (0.016) (0.005)
Sani a ion access (dummy) 0.050* -0.048*** -0.008**
(0.025) (0.011) (0.003)
Wa e access ×Sani a ion access -0.029 0.000 0.009
(0.029) (0.016) (0.005)
R2-squa ed 0.063 0.069 0.018
Mean o ou come 0.834 0.116 0.027
N 37964 67248 67248
Coun ies 19 17 17
Con ols
Coun y- egion-yea FE Yes Yes Yes
Child en le el Yes Yes Yes
Mo he le el Yes Yes Yes
Household cha ac e is ics Yes Yes Yes
No es: This able p esen s esul s om a linea p obabili y model o he co -
ela ion be ween child de elopmen and heal h indica o s and access o wa e
and sani a ion in as uc u e including in e ac ions be ween wa e and san-
i a ion. The analysis sample is composed om mul iple su eys included in
he Mul iple Indica o Clus e Su eys (MICS). The coun ies ha ha e been
included in his analysis and hei espec i e sample sizes can be ound in Ap-
pendix Table A1. The analysis con ols o ixed e ec s a he coun y- egion-
yea le el. Addi ional con ols a he child le el include age (in mon hs), bi h
coho , and sex. Mo he -le el con ols accoun o he mo he ’s educa ion
le el. Household cha ac e is ics con olled o a e u ban esidency s a us,
access o elec ici y, o al household size, and he numbe o household mem-
be s who a e unde i e yea s o age. Va iable de ini ions a e a ailable in
Appendix Table A2 and summa y s a is ics in Appendix Tables A3 and A4.
Su ey clus e ed s anda d e o s a e epo ed in pa en heses (*** p<0.01, **
p<0.05, * p<0.1).
49
Table A10: Access o in as uc u e and he ela ionship wi h de elopmen al and heal h ou comes: mi iga ion beha io s
Dependen a iable ECDI S un ed Unde weigh ECDI S un ed Unde weigh
(dummy) (dummy) (dummy) (dummy) (dummy) (dummy)
(1) (2) (3) (4) (5) (6)
Wa e access (dummy) 0.055* -0.020 -0.008 -0.008 -0.053 -0.020
(0.028) (0.019) (0.005) (0.067) (0.033) (0.015)
Sani a ion access (dummy) 0.071*** -0.055*** -0.008* -0.042 -0.053 -0.015
(0.024) (0.015) (0.004) (0.057) (0.039) (0.016)
Wa e access ×Sani a ion access -0.042 0.007 0.008 0.079 0.017 0.010
(0.029) (0.018) (0.005) (0.072) (0.043) (0.019)
T ea s wa e (dummy) 0.077* -0.029 0.003
(0.038) (0.017) (0.010)
Wa e access ×T ea s wa e -0.038 0.040 -0.004
(0.046) (0.026) (0.013)
Sani a ion access ×T ea s wa e -0.066* 0.030 -0.001
(0.036) (0.019) (0.010)
Wa e access ×Sani a ion access ×T ea s wa e 0.040 -0.030 0.003
(0.049) (0.025) (0.013)
Cleanse o wash hands (dummy) 0.011 -0.049 -0.018
(0.035) (0.029) (0.018)
Wa e access ×Cleanse o wash hands 0.042 0.051* 0.015
(0.048) (0.024) (0.020)
Sani a ion access ×Cleanse o wash hands 0.082* 0.017 0.011
(0.045) (0.041) (0.020)
Wa e access ×Sani a ion access ×Cleanse o wash hands -0.103* -0.026 -0.004
(0.057) (0.045) (0.023)
R2-squa ed 0.064 0.069 0.017 0.065 0.067 0.017
Mean o ou come 0.834 0.116 0.027 0.834 0.114 0.026
N 37907 67171 67171 31466 59141 59141
Coun ies 19 17 17 17 14 14
Con ols
Coun y- egion-yea FE Yes Yes Yes Yes Yes Yes
Child en le el Yes Yes Yes Yes Yes Yes
Mo he le el Yes Yes Yes Yes Yes Yes
Household cha ac e is ics Yes Yes Yes Yes Yes Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween child de elopmen and heal h indica o s and access o wa e and sani a ion
in as uc u e including in e ac ions wi h mi iga ion beha io s such as ea ing wa e o sa e consump ion and ha ing hands cleanse a ailable. The analysis sample is
composed om mul iple su eys included in he Mul iple Indica o Clus e Su eys (MICS). The coun ies ha ha e been included in his analysis and hei espec i e
sample sizes can be ound in Appendix Table A1. The analysis con ols o ixed e ec s a he coun y- egion-yea le el. Addi ional con ols a he child le el include age
(in mon hs), bi h coho , and sex. Mo he -le el con ols accoun o he mo he ’s educa ion le el. Household cha ac e is ics con olled o a e u ban esidency s a us,
access o elec ici y, o al household size, and he numbe o household membe s who a e unde i e yea s o age. Va iable de ini ions a e a ailable in Appendix Table A2
and summa y s a is ics in Appendix Tables A3 and A4. Su ey clus e ed s anda d e o s a e epo ed in pa en heses (*** p<0.01, ** p<0.05, * p<0.1).
Table A11: Access o in as uc u e and he ela ionship wi h de elopmen al and heal h ou comes:
decomposing wa e access
Dependen a iable ECDI S un ed Unde weigh
(dummy) (dummy) (dummy)
(1) (2) (3)
Piped wa e sou ces (base ca ego y)
Tubewell o bo ehole -0.051** 0.010 0.000
(0.021) (0.010) (0.005)
P o ec ed well/sp ing -0.011 0.014 -0.002
(0.007) (0.009) (0.002)
Unp o ec ed sou ces -0.016 0.000 0.002
(0.012) (0.010) (0.003)
T anspo ed wa e sou ces -0.021 -0.001 0.012**
(0.013) (0.017) (0.004)
Packaged wa e 0.009** -0.036*** -0.001
(0.004) (0.009) (0.003)
Su ace wa e -0.053* 0.030** 0.003
(0.029) (0.012) (0.005)
R2-squa ed 0.064 0.069 0.017
Mean o ou come 0.834 0.116 0.027
N 37559 66279 66279
Coun ies 19 17 17
Con ols
Coun y- egion-yea FE Yes Yes Yes
Child en le el Yes Yes Yes
Mo he le el Yes Yes Yes
Household cha ac e is ics Yes Yes Yes
No es: This able p esen s esul s om a linea p obabili y model o he co -
ela ion be ween child de elopmen and heal h indica o s and access o wa e
when decomposing he main wa e access bina y indica o in sub-ca ego ies.
The analysis sample is composed om mul iple su eys included in he Mul i-
ple Indica o Clus e Su eys (MICS). The coun ies ha ha e been included
in his analysis and hei espec i e sample sizes can be ound in Appendix
Table A1. The analysis con ols o ixed e ec s a he coun y- egion-yea
le el. Addi ional con ols a he child le el include age (in mon hs), bi h
coho , and sex. Mo he -le el con ols accoun o he mo he ’s educa ion
le el. Household cha ac e is ics con olled o a e u ban esidency s a us,
access o elec ici y, o al household size, and he numbe o household mem-
be s who a e unde i e yea s o age. Va iable de ini ions a e a ailable in
Appendix Table A2 and summa y s a is ics in Appendix Tables A3 and A4.
Su ey clus e ed s anda d e o s a e epo ed in pa en heses (*** p<0.01, **
p<0.05, * p<0.1).
51
Table A12: Access o in as uc u e and he ela ionship wi h de elopmen al and heal h ou comes:
decomposing sani a ion access
Dependen a iable ECDI S un ed Unde weigh
(dummy) (dummy) (dummy)
(1) (2) (3)
Flush sys ems (base ca ego y)
Hygienic non- lush -0.025*** 0.026*** -0.002
(0.007) (0.007) (0.002)
Unsani a y sys ems -0.039*** 0.048*** -0.001
(0.011) (0.007) (0.004)
No acili ies -0.043*** 0.076*** 0.002
(0.013) (0.013) (0.003)
R2-squa ed 0.063 0.070 0.018
Mean o ou come 0.834 0.116 0.027
N 37795 66928 66928
Coun ies 19 17 17
Con ols
Coun y- egion-yea FE Yes Yes Yes
Child en le el Yes Yes Yes
Mo he le el Yes Yes Yes
Household cha ac e is ics Yes Yes Yes
No es: This able p esen s esul s om a linea p obabili y model o he
co ela ion be ween child de elopmen and heal h indica o s and access o
sani a ion when decomposing he main sani a ion access bina y indica o
in sub-ca ego ies. The analysis sample is composed om mul iple su eys
included in he Mul iple Indica o Clus e Su eys (MICS). The coun ies
ha ha e been included in his analysis and hei espec i e sample sizes can
be ound in Appendix Table A1. The analysis con ols o ixed e ec s a
he coun y- egion-yea le el. Addi ional con ols a he child le el include
age (in mon hs), bi h coho , and sex. Mo he -le el con ols accoun o he
mo he ’s educa ion le el. Household cha ac e is ics con olled o a e u ban
esidency s a us, access o elec ici y, o al household size, and he numbe o
household membe s who a e unde i e yea s o age. Va iable de ini ions a e
a ailable in Appendix Table A2 and summa y s a is ics in Appendix Tables
A3 and A4. Su ey clus e ed s anda d e o s a e epo ed in pa en heses
(*** p<0.01, ** p<0.05, * p<0.1).
52
Table A19: Access o was e managemen in as uc u e and he ela ionship wi h heal h ou comes:
lidded con aine s as a mi iga ion ac o
Dependen a iable S un ed Unde weigh
(dummy) (dummy)
(1) (2)
Ga bage collec ion (dummy) -0.003 -0.003
(0.008) (0.006)
Ga bage collec ion ×Lidless ( s. lidded) -0.040 0.008
(0.021) (0.007)
Ga bage collec ion ×Bags ( s. lidded) -0.016 0.003
(0.010) (0.007)
Ga bage collec ion ×O he ( s. lidded) -0.036* -0.019
(0.015) (0.011)
R2-squa ed 0.170 0.038
N 42199 42199
Coun ies 5 5
Dep. Va . Mean 0.269 0.048
Con ols
Coun y- egion-yea FE Yes Yes
Child en le el Yes Yes
Mo he le el Yes Yes
Household cha ac e is ics Yes Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween
child heal h indica o s and access o was e managemen in as uc u e including in e ac ions
wi h ga bage con aine ypes p io o collec ion. The analysis sample is composed om su eys
included in he Demog aphic and Heal h Su eys (DHS). The analysis con ols o ixed e ec s
a he coun y- egion-yea le el. Addi ional con ols a he child le el include age (in mon hs),
bi h coho , and sex. Mo he -le el con ols ake in o accoun he mo he ’s educa ion le el and
whe he she is he head o he household. Household cha ac e is ics include u ban esidency
s a us, access o elec ici y, wa e , sewage, o al household size, and numbe o membe s unde
i e yea s old. Va iable de ini ions a e a ailable in Appendix Table A2 and summa y s a is ics in
Appendix Table A5. Su ey clus e ed s anda d e o a e epo ed in pa en heses (*** p<0.01,
** p<0.05, * p<0.1).
59
Table A20: Access o was e managemen in as uc u e and he ela ionship wi h heal h ou comes:
heal h channels
Dependen a iable S un ed Unde weigh
(dummy) (dummy)
(1) (2)
Ga bage collec ion (dummy) -0.024*** 0.003*
(0.004) (0.002)
Ga bage collec ion ×Dia hea (dummy) 0.011* -0.010*
(0.006) (0.005)
Ga bage collec ion ×Cough (dummy) -0.004 -0.008**
(0.007) (0.003)
Ga bage collec ion ×Fe e (dummy) -0.007 -0.007**
(0.005) (0.003)
R2-squa ed 0.172 0.049
N 110647 111124
Coun ies 13 13
Dep. Va . Mean 0.241 0.057
Con ols
Coun y- egion-yea FE Yes Yes
Child en le el Yes Yes
Mo he le el Yes Yes
Household cha ac e is ics Yes Yes
No es: This able p esen s esul s om a linea p obabili y model o he co ela ion be ween
child heal h indica o s and access o was e managemen in as uc u e including in e ac ions
wi h ecen child illness episodes such as dia hea, cough, and e e . The analysis sample is
composed om su eys included in he Demog aphic and Heal h Su eys (DHS). The analysis
con ols o ixed e ec s a he coun y- egion-yea le el. Addi ional con ols a he child le el
include age (in mon hs), bi h coho , and sex. Mo he -le el con ols ake in o accoun he
mo he ’s educa ion le el and whe he she is he head o he household. Household cha ac e is ics
include u ban esidency s a us, access o elec ici y, wa e , sewage, o al household size, and
numbe o membe s unde i e yea s old. Va iable de ini ions a e a ailable in Appendix Table
A2 and summa y s a is ics in Appendix Table A5. Su ey clus e ed s anda d e o s a e epo ed
in pa en heses (*** p<0.01, ** p<0.05, * p<0.1).
60
